Menu
CS

Search

Persons

prof. Ing. Vladimír Havlena, CSc.

All publications

Desensitized Extended Kalman Filter with Stochastic Approach to Sensitivity Reduction and Adaptive Weights

  • Authors: Tabaček, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 2022 25th International Conference on Information Fusion (FUSION). Piscataway: IEEE, 2022. ISBN 978-1-7377497-2-1.
  • Year: 2022
  • DOI: 10.23919/FUSION49751.2022.9841381
  • Link: https://doi.org/10.23919/FUSION49751.2022.9841381
  • Department: Department of Control Engineering
  • Annotation:
    The desensitized Kalman filter can robustly estimate the state of a system with uncertain parameters without knowledge about uncertainty type. In this paper, the desensitized Kalman filter for nonlinear systems is derived using Taylor series expansion and a stochastic approach to reduce estimation error sensitivity to uncertain parameters. Adaptively normalized weights tune the trade-off between the minimum uncertainty sensitivity and minimum mean square error. Among the main benefits of the algorithm are intuitive tuning concerning uncertainty and a form resembling the classical Riccati equation. The comparison to other robust state-of-the-art algorithms is discussed based on a numerical example.

Fault detection for buildings using uncertain parameters and interacting multiple-model method

  • Authors: Tabaček, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Lecture Notes in Control and Information Sciences - Proceedings. Wien: Springer, 2022. ISSN 2522-5383. ISBN 978-3-030-85317-4.
  • Year: 2022
  • DOI: 10.1007/978-3-030-85318-1_77
  • Link: https://doi.org/10.1007/978-3-030-85318-1_77
  • Department: Department of Control Engineering
  • Annotation:
    Model-based fault detection and diagnosis (FDD) systems for buildings are very demanding on the solution set up eort. One rea- son is the requirement to use a high-delity model which must be created for each building separately. The FDD approach that reduces this bur- den is proposed in this paper. Proposed FDD algorithms are based on the interacting multiple-model (IMM) method and Kalman ltering for systems with uncertain parameters. The uncertain parameters in models enable to use "average" zone models rather than high-delity models, which simplies real applications. Detection of single and multiple faults is demonstrated on an example where faults, that might cause ine- ciency of control, are detected. Results show that the performance of the proposed algorithms with average zone models is comparable to the performance of the conventional IMM with accurate models.

Reduction of prediction error sensitivity to parameters in Kalman filter

  • Authors: Tabaček, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS. 2022, 359(3), 1303-1326. ISSN 0016-0032.
  • Year: 2022
  • DOI: 10.1016/j.jfranklin.2021.12.019
  • Link: https://doi.org/10.1016/j.jfranklin.2021.12.019
  • Department: Department of Control Engineering
  • Annotation:
    The desensitized Kalman filter is a practical and intuitive robust filtering method. However, a thorough analysis of its stability and impact of assumptions is missing. This paper expands the theory of desensitized Kalman filtering by proposing a stochastic approach to reduce estimation error sensitivity to parameters. The novel approach leads to the exact desensitized Kalman filter that does not neglect the gain sensitivity to a parameter. The suboptimal form equivalent to the original desensitized Kalman filter in a special form is proposed. The stability analysis and the definition of stability conditions are possible due to the proposed form that can be interpreted as the Kalman filter with correlated process and measurement noise with time-variant statistics. Furthermore, adaptive normalization of objectives is introduced, which improves the desensitizing performance.

Mixed Mesh Finite Volume Method for 1D Hyperbolic Systems with Application to Plug-flow Heat Exchangers

  • DOI: 10.3390/math9202609
  • Link: https://doi.org/10.3390/math9202609
  • Department: Department of Control Engineering
  • Annotation:
    We present a finite volume method formulated on a mixed Eulerian-Lagrangian mesh for highly advective 1D hyperbolic systems altogether with its application to plug-flow heat exchanger modeling/simulation. Advection of sharp moving fronts is an important problem in fluid dynamics, and even a simple transport equation cannot be solved precisely by having a finite number of nodes/elements/volumes. Finite volume methods are known to introduce numerical diffusion, and there exist a wide variety of schemes to minimize its occurrence; the most recent being adaptive grid methods such as moving mesh methods or adaptive mesh refinement methods. We present a solution method for a class of hyperbolic systems with one nonzero time-dependent characteristic velocity. This property allows us to rigorously define a finite volume method on a grid that is continuously moving by the characteristic velocity (Lagrangian grid) along a static Eulerian grid. The advective flux of the flowing field is, by this approach, removed from cell-to-cell interactions, and the ability to advect sharp fronts is therefore enhanced. The price to pay is a fixed velocity-dependent time sampling and a time delay in the solution. For these reasons, the method is best suited for systems with a dominating advection component. We illustrate the method’s properties on an illustrative advection-decay equation example and a 1D plug flow heat exchanger. Such heat exchanger model can then serve as a convection-accurate dynamic model in estimation and control algorithms for which it was developed.

On the Quadratic Programming Solution for Model Predictive Control with Move Blocking

  • Authors: Otta, P., Šantin, O., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 23rd International Conference on Process Control. Piscataway: IEEE, 2021. p. 49-54. ISBN 978-1-6654-0330-6.
  • Year: 2021
  • DOI: 10.1109/PC52310.2021.9447499
  • Link: https://doi.org/10.1109/PC52310.2021.9447499
  • Department: Department of Control Engineering
  • Annotation:
    Model Predictive Control (MPC) is a popular optimization-based control technique. MPC is usually formulated as sparse or dense Quadratic Programming (QP). This paper reviews two well-known methods: state condensing and move blocking, and brings them together. Their combination results in general QP that serves arbitrarily sparse (or dense) QP for MPC with move blocking. The proposed QP can be solved by a specialized solver capable of exploiting a problem's sparsity structure. A numerical example illustrates connections to computational and memory requirements.

Range control MPC with application to Vapor Compression Cycles

  • DOI: 10.1016/j.conengprac.2020.104309
  • Link: https://doi.org/10.1016/j.conengprac.2020.104309
  • Department: Department of Control Engineering
  • Annotation:
    A range control MPC formulation is presented for control of Vapor Compression Cycles (VCC). Variation in system dynamics is handled through application of range control which is a generalization of a tracking problem whereby the set-point for the system output is replaced by a funnel (time-varying upper and lower bounds on the system output) which is implemented as a set of soft constraints. Vapor cycle is optimized to maximize device efficiency while ensuring feasibility of operation. Experimental validation of the control design, embedded in a heat-pump controller, is presented.

Gaussian Process Based Model-free Control with Q-Learning

  • DOI: 10.1016/j.ifacol.2019.09.147
  • Link: https://doi.org/10.1016/j.ifacol.2019.09.147
  • Department: Department of Control Engineering
  • Annotation:
    The aim of this paper is to demonstrate a new algorithm for Machine Learning (ML) based on Gaussian Process Regression (GPR) and how it can be used as a practical control design technique. An optimized control law for a nonlinear process is found directly by training the algorithm on noisy data collected from the process when controlled by a suboptimal controller. A simplified nonlinear Fan Coil Unit (FCU) model is used as an example for which the fan speed control is designed using the off-policy Q -learning algorithm. Additionally, the algorithm properties are discussed, i.e. learning process robustness, Gaussian Process (GP) kernel functions choice. The simulation results are compared to a simple PI design based on a linearized model. (C) 2019, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

Newton projection with proportioning using iterative linear algebra for model predictive control with long prediction horizon

  • Authors: Otta, P., Burant, J., Šantin, O., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Optimization Methods and Software. 2019, 34(5), 1075-1098. ISSN 1055-6788.
  • Year: 2019
  • DOI: 10.1080/10556788.2019.1571588
  • Link: https://doi.org/10.1080/10556788.2019.1571588
  • Department: Department of Control Engineering
  • Annotation:
    This paper presents an algorithm to solve a sparse Quadratic Programming (QP) problem. The QP problem is suitable for Model Predictive Control (MPC) applications in particular. MPC is a modern multivariable control method which requires the solution to a quadratic programming problem at each sampling instant. The proposed algorithm is an active-set based strategy which uses the proportioning test for the selection of the active-set reduction and expansion while utilizing the sparse nature of the problem by the preconditioned MINRES algorithm to solve the face problem. Numerical experiments illustrate the performance of the algorithm, and the results are compared with the state-of-the-art solvers.

Desensitized Filtering for Systems with Uncertain Parameters and Noise Correlation

  • Authors: Tabaček, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: The 26th Mediterranean Conference on Control and Automation. Piscataway, NJ: IEEE, 2018. p. 794-799. ISSN 2325-369X. ISBN 978-1-5386-7890-9.
  • Year: 2018
  • DOI: 10.1109/MED.2018.8443049
  • Link: https://doi.org/10.1109/MED.2018.8443049
  • Department: Department of Control Engineering
  • Annotation:
    This paper introduces estimation algorithms for systems with uncertain parameters and correlated noises. The algorithms are derived using the standard Kalman filter for correlated noises and the desensitized filtering technique for systems with uncertain parameters. A general algorithm and its special case are proposed. The latter updates statistics with explicit expressions, which makes it simpler and faster. The extended forms of the algorithms, which can be used for nonlinear systems, are also introduced. The developed algorithm is tested on an example, where the importance of the noise correlation information is shown.

Schmidt-Kalman Filters for Systems with Uncertain Parameters and Asynchronous Sampling

  • Authors: Tabaček, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 18th International Conference on Control, Automation and Systems. Wonmi-gu, Gyeonggi-do: Institute of Control, Robotics and Systems (ICROS), 2018. p. 357-362. ISSN 2093-7121. ISBN 978-89-93215-15-1.
  • Year: 2018
  • Department: Department of Control Engineering
  • Annotation:
    This paper introduces estimation algorithms for systems with uncertain parameters and asynchronous sampling. The algorithms are created by merging the Schmidt-Kalman filter (SKF) for systems with uncertain parameters and the conventional Kalman filter for systems with correlated noises. The system descriptions obtained by different discretization approaches are analyzed and used to develop the equivalent of the SKF. Then the SKF for systems with asynchronous sampling is developed by applying the SKF or its equivalent on the part of sampling period where the process and measurement noises are correlated. The accuracy of the novel filters is tested on a simple example.

Control-oriented models for turbocharged engines

  • Authors: Pachner, D., Tabaček, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Turbochargers and turbocharging: Advancements, applications and research. Hauppauge NY: Nova Science Publisher, Inc., 2017. p. 435-452. ISBN 978-1-5361-2239-8.
  • Year: 2017
  • Department: Department of Control Engineering
  • Annotation:
    This chapter studies numerical aspects of lumped parameter models of gas exchange process in the turbocharged engine. It proposes an efficient method implementing a variable order model based on rational approximations of the flow characteristics. Such models can be simulated and linearized efficiently. The approach is demonstrated comparing the proposed structure with a standard full order stiff model.

Proportioning with second-order information for model predictive control

  • Authors: Šantin, O., Jarošová, M., prof. Ing. Vladimír Havlena, CSc., Dostál, Z.
  • Publication: Optimization Methods and Software. 2017, 32(3), 436-454. ISSN 1055-6788.
  • Year: 2017
  • DOI: 10.1080/10556788.2016.1213840
  • Link: https://doi.org/10.1080/10556788.2016.1213840
  • Department: Department of Control Engineering
  • Annotation:
    We propose an algorithm for the effective solution of quadratic programming (QP) problems arising from model predictive control (MPC). MPC is a modern multivariable control method which gives the solution for a QP problem at each sample instant. Our algorithm combines the active-set strategy with the proportioning test to decide when to leave the actual active set. For the minimization in the face, we use a direct solver implemented by the Cholesky factors updates. The performance of the algorithm is illustrated by numerical experiments, and the results are compared with the state-of-the-art solvers on benchmarks from MPC.

Robust numerical approach to steady-state calibration of mean-value models

  • DOI: 10.1016/j.conengprac.2016.04.009
  • Link: https://doi.org/10.1016/j.conengprac.2016.04.009
  • Department: Department of Control Engineering
  • Annotation:
    A numerically robust approach to steady-state calibration of nonlinear dynamic models is presented. The approach is based on explicit formulation of the constraints on validity of internal model signals by set of inequalities. The constrained optimization with feasible iterates guarantees that the model will never be evaluated with invalid internal signals. This overcomes numerical difficulties often encountered when dealing with highly nonlinear models. Because the approach uses a large number of slack variables, distributed least squares algorithm is proposed. The robustness of this approach is demonstrated on a steady-state calibration of turbocharged diesel engine model starting from grossly inaccurate initial estimates.

Convection Oriented Heat Exchanger Model

  • Authors: Dostál, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 2016 12th IEEE International Conference on Control and Automation (ICCA). Piscataway, NJ: IEEE, 2016. pp. 347-352. ISSN 1948-3449. ISBN 978-1-5090-1738-6.
  • Year: 2016
  • DOI: 10.1109/ICCA.2016.7505301
  • Link: https://doi.org/10.1109/ICCA.2016.7505301
  • Department: Department of Control Engineering
  • Annotation:
    The paper focuses on an accurate modeling of an outlet water temperature in a single phase water-to-air heat exchanger subject to dynamic changes on any inlet. Particularly, a convection character of the water stream is taken into account such that a finite, mass flow driven, temperature/information propagation speed through the heat exchanger is achieved. Finite difference approximation methods demonstrate numerical diffusion resulting in an infinite information propagation speed, therefore a moving water element approach is adopted and a convection oriented heat exchanger model is derived. Hybrid representation as well as a simplified discrete time model of the heat exchanger are given. An example showing improved convection behavior closes the paper.

Convection Oriented Heat Exchanger Model - Identification

  • Authors: Dostál, J., Prajzner, V., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: CLIMA 2016 - proceedings of the 12th REHVA World Congress: volume 9. Aalborg: Aalborg University, 2016. ISBN 87-91606-34-9.
  • Year: 2016
  • Department: Department of Control Engineering
  • Annotation:
    A convection oriented heat exchanger model, proposed in [1], features the ability to advect sharp moving fronts in the water stream as opposed to a traditional finite difference numerical solution scheme. The model is capable of predicting heat exchanger output temperatures subject to simultaneous dynamic input changes for a wide range of flow rates including a zero flow. This paper extends the convection oriented heat exchanger model proposed in [1] by an identification of heat exchange coefficient functions and presents a successful validation of the method by experimental data.

Distributed MPC with parametric coordination

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc., Pekař, J.
  • Publication: 2016 American Control Conference ACC. Piscataway, NJ: IEEE, 2016. p. 6253-6258. ISSN 0743-1619. ISBN 978-1-4673-8682-1.
  • Year: 2016
  • DOI: 10.1109/ACC.2016.7526652
  • Link: https://doi.org/10.1109/ACC.2016.7526652
  • Department: Department of Control Engineering
  • Annotation:
    The paper presents an effective coordination scheme for a distributed optimization based on dual decomposition. The targeted class of optimization problems are strictly convex quadratic functions with linear constraints, where the dual function with coupling equality constraints in Lagrangian is continuous piecewise quadratic. The coordination is based on multi-parametric programming - the subproblem solvers return their solution on a polyhedron around a given Lagrange multiplier value. Centralized coordinator constructs the gradient and Hessian of a dual function and reaches exact consensus in a finite number of iterations, while only some subproblems are queried for a new solution in each iteration. The algorithm is applied to the distributed model predictive control. The efficiency, in terms of optimization time and the number of iterations, is demonstrated on the distributed model predictive control of the Barcelona water distribution network

Measured-State Driven Warm-Start Strategy for Linear MPC

  • Authors: Otta, P., Šantin, O., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the European Control Conference. Brussels: EUCA, 2015. p. 3132-3136. ISBN 978-3-9524269-3-7.
  • Year: 2015
  • DOI: 10.1109/ECC.2015.7331015
  • Link: https://doi.org/10.1109/ECC.2015.7331015
  • Department: Department of Control Engineering
  • Annotation:
    Model Predictive Control (MPC) is an optimization-based control technique which involves solving an optimization problem in every sampling instant. As a consequence, MPC is very computationally demanding, compared to traditional control techniques. Thus, much effort in academia and also in industry is aimed at finding a way to decrease the computation time. One approach is to find such a starting point for the iterative solver that it finds the optimum in less iterations - this technique is called warm-start or hot-start. This paper brings attention back to the common warm-start technique as well as an alternative - realized via the Linear Quadratic Regulator (LQR) static feedback. The paper proposes a natural combination of both of these warm-start techniques to achieve improved results.

A Quasi-1D Model of Biomass Co-Firing in a Circulating Fluidized Bed Boiler

  • Authors: Beneš, M., Strachota, P., Máca, R., prof. Ing. Vladimír Havlena, CSc., Mach, J.
  • Publication: Finite Volumes for Complex Applications VII, Elliptic, Parabolic and Hyperbolic Problems. Berlin: Springer Science+Business Media, 2014. p. 791-799. Springer Proceedings in Mathematics and Statistics. ISSN 2194-1009. ISBN 978-3-319-05590-9.
  • Year: 2014
  • DOI: 10.1007/978-3-319-05591-6_79
  • Link: https://doi.org/10.1007/978-3-319-05591-6_79
  • Department: Department of Control Engineering
  • Annotation:
    We introduce an outline of the mathematical model of combustion in circulating fluidized bed boilers. The model is concerned with multiphase flow of flue gas, bed material, and two types of fuels (coal and biomass) that can be co-fired in the furnace. It further considers phase interaction resulting in particle attrition, devolatilization and burnout of fuel particles, and energy balance between heat production and consumption (radiative and convective transfer to walls). Numerical solution by means of the finite volume method together with a Runge-Kutta class time integration scheme is mentioned only briefly as the used methods are generic and well documented elsewhere. Some representative results are also presented.

Maximum Likelihood Estimation of LTI Continuous-Time Grey-box Models

  • Authors: Řehoř, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of The 19th World Congress of the International Federation of Automatic Control. Pretoria: IFAC, 2014. pp. 3739-3744. ISSN 1474-6670. ISBN 978-3-902823-62-5.
  • Year: 2014
  • Department: Department of Control Engineering
  • Annotation:
    Grey-box models give us a welcomed opportunity to combine our prior knowledge with experimental data when a system is being identified. This benefit is redeemed by an unattractive non-convex and nonlinear optimisation problem that ensue from the parameter estimation. The article shows an efficient method how to speed up the arising iterative optimisation algorithm in the case of continuous, time-invariant, grey-box models. The method is based on a presented output-to-parameter sensitivity computation algorithm.

Modeling, optimization and analysis of hydronic networks with decentralized pumping

  • Authors: Dostál, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 2014 CACS International Automatic Control Conference. Peking: Chinese Automatic Control Society, 2014. p. 269-274. ISBN 978-1-4799-4584-9.
  • Year: 2014
  • Department: Department of Control Engineering
  • Annotation:
    A new concept of decentralized pumping in residential heating and cooling systems has been proposed in recent years and it is generally declared to save pumping energy. A set of tools has been derived in order to verify this proposition. A graph modeling method altogether with a network flow solver and pumping energy optimization solver are presented in the paper. The center-point, however, lies in a power consumption analysis of throttling (supply oriented, central) systems and decentralized pumping (demand oriented, distributed) systems. A pumping strategy comparison for a sample residential hydronic network illustrates the modeling approach and power optimization results.

Tailored QP Algorithm for Predictive Control with Dynamics Penalty

  • Authors: Otta, P., Šantin, O., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 22nd Mediterranean Conference on Control & Automation. Piscataway: IEEE, 2014. p. 384-389. ISBN 978-1-4799-5900-6.
  • Year: 2014
  • DOI: 10.1109/MED.2014.6961402
  • Link: https://doi.org/10.1109/MED.2014.6961402
  • Department: Department of Control Engineering
  • Annotation:
    In order to reduce the computational complexity of solving Quadratic Programming (QP), related to linear Model Predictive Control (MPC), a new approximated formulation of the QP with simple bounds is introduced in this paper. This formulation is based on the idea not to consider model dynamics as a hard constraint but rather modify the objective function of MPC by penalty to capture the violation of model dynamics.The system dynamics is usually uncertain and then it does not make sense to design the control law based on the exact model. Furthermore, the specific sparse structure of the approximated simple bounded QP formulation of the MPC problem is exploited in the new type of combined gradient/Newton step projection algorithm with linear complexity of each iteration with respect to prediction horizon. It is shown by examples that the proposed method is faster on tested problem than other state-of-the-art solvers while retaining a high performance level.

MPC-based approximate dual controller by information matrix maximization

  • Authors: Rathouský, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: International Journal of Adaptive Control and Signal Processing. 2013, 27(11), 974-999. ISSN 0890-6327.
  • Year: 2013
  • DOI: 10.1002/acs.2370
  • Link: https://doi.org/10.1002/acs.2370
  • Department: Department of Control Engineering
  • Annotation:
    This paper proposes a method to approximate a dual controller by a omputationally feasible algorithm. Dual control that optimally solves the problem of simultaneous control and identification of a system with uncertain parameters is known to be both analytically and computationally unsolvable. This paper proposes a multiple-step active control algorithm that gives a suboptimal but tractable solution to the original dual kontrol problem. The algorithm is based on model predictive control (MPC) and approximates persistent systém excitation in terms of the increase of the lowest eigenvalue of the parameter estimate information matrix.

Noise covariance estimation for Kalman filter tuning using Bayesian approach and Monte Carlo

  • Authors: Matisko, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: International Journal of Adaptive Control and Signal Processing. 2013, 27(11), 957-973. ISSN 0890-6327.
  • Year: 2013
  • DOI: 10.1002/acs.2369
  • Link: https://doi.org/10.1002/acs.2369
  • Department: Department of Control Engineering
  • Annotation:
    Linear time-invariant systems play significant role in the control field. A number of methods have been published for identification of the deterministic part of a process. However, identification of the stochastic part has had much less attention. This paper deals with estimation of covariance matrices of the noise entering a linear system. The process and measurement noise covariance matrices are tuning parameters of the Kalman filter and they affect quality of the state estimation. The noise covariance matrices are generally not known and their estimation from the measured data is a challenging task. This paper introduces a method for estimation of the noise covariance matrices using Bayesian approach along with Monte Carlo numerical methods. Performance of the approach is tested on various systems and noise properties. The second part of the paper compares Monte Carlo approach to the recently published methods. The speed of convergence is compared to the Cramér-Rao bounds.

Robust Numerical Approach to Mean-Value Modeling of Internal Combustion Engines

  • Authors: Beňo, R., Pachner, D., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Preprints of the 7th IFAC Symposium on Advances in Automotive. Laxenburg: IFAC, 2013, pp. 564-569. ISSN 1474-6670. ISBN 978-3-902823-43-4.
  • Year: 2013
  • DOI: 10.3182/20130904-4-JP-2042.00078
  • Link: https://doi.org/10.3182/20130904-4-JP-2042.00078
  • Department: Department of Control Engineering
  • Annotation:
    A numerically robust approach to the mean value modeling of turbocharged internal combustion engines (ICE) is presented. The approach is based on model regularization on a polyhedron in a high dimensional vector space. Distributed programming techniques can be used for the regularized model identification and simulation. The robustness of this approach is demonstrated on steady-state calibration of a heavy duty diesel engine model starting from grossly inaccurate initial estimates.

Robust optimization on receding horizon of processes with storages and periodic production and consumption contracts

  • Authors: Baramov, L., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the European Control Conference. Zurich: European Control Association, 2013, pp. 4448-4453. ISBN 978-3-033-03962-9.
  • Year: 2013
  • Department: Department of Control Engineering
  • Annotation:
    This paper presents a real-time optimizer for processes with material storages and periodic targets on integrals of optimized variables. A problem of robust target attainment with respect to uncertainties in future disturbance trajectory estimates was formulated. This robustness is achieved by means of soft constraints forcing future trajectories to a time-varying set from which reaching the target is feasible. An algorithm for computing this set is outlined, which is tractable for problems of medium size. The application area for this optimizer is assumed to be in (but not limited to) heat and power plant optimization with multiple fuels with periodic contracts on electric energy delivered and fuel gas consumed during the contract period.

Structured Model Order Reduction of Parallel Models in Feedback

  • Authors: Trnka, P., Sturk, Ch., Sandberg, H., prof. Ing. Vladimír Havlena, CSc., Řehoř, J.
  • Publication: IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY. 2013, 21(3), 739-752. ISSN 1063-6536.
  • Year: 2013
  • DOI: 10.1109/TCST.2012.2192735
  • Link: https://doi.org/10.1109/TCST.2012.2192735
  • Department: Department of Control Engineering
  • Annotation:
    Parallel working units in closed-loop operation are frequently encountered in industrial applications of advanced process control (boilers, turbines, chemical reactors, etc.). Control strategies typically require different low-order models for each configuration of parallel units. These different models are usually obtained by heuristics applied to the parallel models. To replace these heuristics, this paper proposes a systematic solution based on structured model order reduction. Two methods are considered, the first has general applicability to stable closed-loop systems, but gives no a priori error bounds; the second linear matrix inequality (LMI)-based method comes with an explicit error bounds, but cannot be applied to general models. However, it is shown that for models composed of cascades of stable subsystems and negative feedbacks of strictly positive real subsystems, the LMIs are always feasible. Both methods are demonstrated on a practical example of a cogeneration power plant with multiple boilers. It is proved that the second LMI-based method can always be applied to general problems with structures similar to the boiler-header systems considered in this paper.

Biofuel co-firing with inferential sensor

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Power Plants and Power Systems Control. Toulouse: Université de Toulouse, 2012, pp. 717-722. ISSN 1474-6670. ISBN 978-3-902823-24-3.
  • Year: 2012
  • DOI: 10.3182/20120902-4-FR-2032.00125
  • Link: https://doi.org/10.3182/20120902-4-FR-2032.00125
  • Department: Department of Control Engineering
  • Annotation:
    Biomass provides large carbon dioxide reduction potential when it replaces coal as the dominant fuel in large scale heat and electricity production. However, the volatility of biomass properties may become a limiting factor to the proportion of biomass used. The paper presents a model based control strategy for multiple fuel co-firing, which enables direct compensation of the variability of fuel properties and stabilizes the heat power using an inferential sensing approach. Kalman filter with significantly increased robustness with respect to model uncertainty is required to make the inferential sensing practically applicable.

Cramér-Rao Bounds for Estimation of Linear System Noise Covariances

  • DOI: 10.5923/j.jmea.20120202.02
  • Link: https://doi.org/10.5923/j.jmea.20120202.02
  • Department: Department of Control Engineering
  • Annotation:
    The performance of Kalman filter depends directly on the noise covariances, which are usually not known and need to be estimated. Several estimation algorithms have been published in past decades, but the measure of estimation quality is missing. The Cramér-Rao bounds represent limitation of quality of parameter estimation that can be obtained from given data. In this article, The Cramér-Rao bounds for noise covariance estimation of linear time-invariant stochastic system will be derived. Two different linear system models will be considered. Further, the performance of earlier published methods will be discussed according to the Cramér-Rao bounds. The analogy between the Cramér-Rao bounds and the Riccati equation will be pointed out.

Enforcing Stability in Steady-State Optimization

  • Authors: Beňo, R., Pachner, D., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 16th IFAC Symposium on System Identification. Laxenburg: IFAC, 2012, pp. 1599-1604. ISSN 1474-6670. ISBN 978-3-902823-06-9.
  • Year: 2012
  • Department: Department of Control Engineering
  • Annotation:
    The article deals with the stability constraint in nonlinear continuous-time dynamic model identification. The identification is formulated as a boundary value problem. Constraining the norm of the terminal sensitivity to the initial condition is used to drive the model state to a stable equilibrium. Solving such boundary value problem on an extending finite time horizon may be numerically more appealing than constraining the eigenvalues of the Jacobian matrix evaluated at the equilibrium point in the state space.

Mathematical modelling of combustion and biofuel co-firing in industrial steam generators

  • Authors: Beneš, M., Oberhuber, T., Strachota, P., Straka, R., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: RIMS Kokyuroku. 2012, B35 141-157. ISSN 1880-2818.
  • Year: 2012
  • Department: Department of Control Engineering
  • Annotation:
    In the contribution, we summarize results of mathematical modelling and numerical simulation of flow, transport, combustion and reaction processes in industrial steam generators powered by the powderized coal combustion with possible partial biofuel additives (biofuel co-firing).

Optimality Tests and Adaptive Kalman Filter

  • Authors: Matisko, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of 16th IFAC Symposium on System Identification. Brusel: IFAC, 2012, pp. 1-6. ISSN 1474-6670. ISBN 978-3-902823-06-9.
  • Year: 2012
  • DOI: 10.3182/20120711-3-BE-2027.00011
  • Link: https://doi.org/10.3182/20120711-3-BE-2027.00011
  • Department: Department of Control Engineering
  • Annotation:
    Kalman filter tuning is based on the process and measurement noise covariances that are often obtained by ad hoc methods. After the filter is tuned, it is necessary to evaluate the quality of the state estimation. In this article, several methods are described for the quality evaluation of the Kalman filter performance. The article includes simulation results evaluating the reliability of the described optimality tests. The sequential test is then used for an adaptive algorithm for a Kalman filter. Further, properties of an autocorrelation function are discussed and several methods for its estimation are compared.

UNSCENTED KALMAN FILTER REVISITED - HERMITE-GAUSS QUADRATURE APPROACH

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of 15th International Conference on Information Fusion. Piscataway: IEEE, 2012, pp. 1-5. ISBN 9780982443842.
  • Year: 2012
  • Department: Department of Control Engineering
  • Annotation:
    Kalman filter is a frequently used tool for linear state estimation due to its simplicity and optimality. It can further be used for fusion of information obtained from multiple sensors. Kalman filtering is also often applied to nonlinear systems. As the direct application of bayesian functional recursion is computationally not feasible, approaches commonly taken use either a local approximation - Extended Kalman Filter based on linearization of the non-linear model - or the global one, as in the case of Particle Filters. An approach to the local approximation is the so called Unscented Kalman Filter. It is based on a set of symmetrically distributed sample points used to parameterise the mean and the covariance. Such filter is computationally simple and no linearization step is required. Another approach to selecting the set of sample points based on decorrelation of multivariable random variables and Hermite-Gauss Quadrature is introduced in this paper. This approachprovides an additional justification of the Unscented Kalman Filter development and provides further options to improve the accuracy of the approximation, particularly for polynomial nonlinearities. A detailed comparison of the two approaches is presented in the paper.

A Note on the Role of the Natural Condition of Control in the Estimation of DSGE Models

  • Department: Department of Control Engineering
  • Annotation:
    This paper is written by authors from the technical and economic fields who are motivated to find a common language and views on the problem of the optimal use of information in model estimation. The center of our interest is the natural condition of control - a common assumption in Bayesian estimation in the technical sciences, and one which may be violated in economic applications. In estimating dynamic stochastic general equilibrum (DSGE) models, typically only a subset of endogenous variables is treated as measured even if additional data sets are available. The natural condition of control dictates the exploitation of all available information, which improves model adaptability and estimate efficiency. We illustrate our points on a basic RBC model.

A Practical Approach to Grey-box Model Identification

  • Authors: Řehoř, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 18th IFAC World Congress, 2011. Bologna: IFAC, 2011, pp. 10776-10781. ISSN 1474-6670. ISBN 978-3-902661-93-7.
  • Year: 2011
  • Department: Department of Control Engineering
  • Annotation:
    Today's computing power can provide a way how to fit physically well founded models to a given experimental data. The combination of user's subjective information and objective data (grey-box modeling) is especially profitable when obtaining sufficiently excited data is impracticable or costly. This is typical in the industrial practice. However a simple prediction error fitting scheme can results in poor estimate of parameters. In this article, a new method how to improve parameters estimate to be relevant in commissioning model predictive control, is proposed. The method is derived on the base of maximum likelihood estimation and demonstrated using a practical application.

Application of Distributed MPC to Barcelona water distribution network

  • Authors: Trnka, P., Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 18th IFAC World Congress, 2011. Bologna: IFAC, 2011. pp. 9139-9144. ISBN 978-3-902661-93-7.
  • Year: 2011
  • DOI: 10.3182/20110828-6-IT-1002.02853
  • Link: https://doi.org/10.3182/20110828-6-IT-1002.02853
  • Department: Department of Control Engineering
  • Annotation:
    The paper presents application of Distributed Model Predictive Control (DMPC) schemes to complex system of Barcelona water distribution network. The dual decomposition of convex optimization problems is well known and has been already adopted to DMPC. However, the application of dual based DMPC to truly large scale systems requires efficient algorithms for consensus iterations. The paper treats DMPC with and without centralized coordinator. The non-centralized coordination is based on Nesterov accelerated gradient method and centralized coordination is based on limited memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method. © 2011 IFAC.

Combined Gradient and Newton Projection Quadratic Programming Solver for MPC

  • Authors: Šantin, O., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 18th IFAC World Congress, 2011. Bologna: IFAC, 2011, pp. 5567-5572. ISSN 1474-6670. ISBN 978-3-902661-93-7. Available from: http://www.ifac-papersonline.net/Detailed/49311.html
  • Year: 2011
  • DOI: 10.3182/20110828-6-IT-1002.02573
  • Link: https://doi.org/10.3182/20110828-6-IT-1002.02573
  • Department: Department of Control Engineering
  • Annotation:
    The objective of this paper is to present an effective on-line solver for simple constrained quadratic programming (QP) which arises in linear model predictive control (MPC) framework. In MPC, the QP is solved at each sampling time, thus a fast solver must be used for short sampling times in real-time applications. The multi-parametric quadratic programming (mp-QP) approach (explicit solution) is impossible to use for larger systems due to the memory limitation. On the other hand, the presented approach is well suitable even for medium scale systems with short sampling time, since it is based on combination of gradient and Newton projection algorithm which is very close to optimum in a very few iterations and the computation of the Newton step is not involved at each iteration.

Combined Partial Conjugate Gradient and Gradient Projection solver for MPC

  • Authors: Šantin, O., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: IEEE Multi-Conference on Systems and Control. Piscataway, NJ: IEEE, 2011. p. 1270-1275. ISSN 1085-1992. ISBN 978-1-4577-1061-2.
  • Year: 2011
  • Department: Department of Control Engineering
  • Annotation:
    The objective of this paper is to present fast algorithm for large scale box constrained quadratic program which arises from linear model predictive control (MPC) with hard limits on the inputs. The presented algorithm uses the combination of the gradient projection method and the partial conjugate gradient method. The special structure of the MPC problem is exploited so that the conjugate gradient method converges in a few number of iterations and the algorithm well suitable for processes with thousands of inputs and small number of outputs is obtained.

Cramér-Rao bounds for estimation of linear system noise covariances

  • Authors: Matisko, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 18th IFAC World Congress, 2011. Bologna: IFAC, 2011, pp. 7791-7796. ISSN 1474-6670. ISBN 978-3-902661-93-7. Available from: http://www.ifac-papersonline.net/Detailed/50045.html
  • Year: 2011
  • DOI: 10.3182/20110828-6-IT-1002.01531
  • Link: https://doi.org/10.3182/20110828-6-IT-1002.01531
  • Department: Department of Control Engineering
  • Annotation:
    The performance of Kalman filter depends directly on the noise covariances, which are usually not known and need to be estimated. Several estimation algorithms have been published in past decades, but the measure of estimation quality is missing. Cramér-Rao bounds represent limitation of quality of parameter estimation that can be obtained from given data. In this article, Cramér-Rao bounds for noise covariance estimation of linear time-invariant stochastic system will be derived. Two different linear system models will be considered. Further, the performance of earlier published methods will be discussed according to Cramér-Rao bounds. The analogy between Cramér-Rao bounds and Riccati equation will be pointed out.

Gradient Projection based Algorithm for Large Scale Real Time Model Predictive Control

  • Authors: Šantin, O., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 2011 Chinese Control and Decision Conference. Shenyang: Northeastern University, 2011. p. 3906-3911. ISBN 978-1-4244-8736-3.
  • Year: 2011
  • Department: Department of Control Engineering
  • Annotation:
    In model predictive control (MPC), the quadratic program (QP) is solved at each sampling time, thus a fast and effective on-line solver must be used for short sampling times. The multi-parametric quadratic programming (mp-QP) (explicit solution) is impossible to use for larger systems due to the memory limitation. The objective of this paper is to present an effective on-line solver for large-scale simple constrained quadratic programming which arises in the MPC framework. The presented algorithm uses the combination of gradient and Newton projection method to obtain super-linear convergent algorithm which is very close to optimum in very few iterations when many constraints are active in optimum and it does not involve the exact computation of the Newton step at each iteration.

MPC-based approximation of dual control by information maximization

  • Authors: Rathouský, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 18th International Conference on Process Control. Bratislava: Slovenská technická univerzita, 2011, pp. 247-252. ISBN 978-80-227-3517-9.
  • Year: 2011
  • Department: Department of Control Engineering
  • Annotation:
    This paper proposes multiple-step active control algorithms based on MPC approach that approximate persistent system excitation in terms of the increase of the lowest eigenvalue of the parameter estimate information matrix. It is shown how the persistent excitation condition is connected with a proposed concept of stability of a system with uncertain parameters. Unlike similar methods, the proposed algorithms predict the information matrix for more than one step of control. The problem is formulated as an MPC problem with an additional constraint on the information matrix. This constraint makes the problem non-convex, thus only locally optima solutions are guaranteed.

Multiple-step active control with dual properties

  • Authors: Rathouský, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 18th IFAC World Congress, 2011. Bologna: IFAC, 2011, pp. 1522-1527. ISSN 1474-6670. ISBN 978-3-902661-93-7.
  • Year: 2011
  • Department: Department of Control Engineering
  • Annotation:
    This paper proposes multiple-step active control algorithms based on MPC approach that approximate persistent system excitation in terms of the increase of the lowest eigenvalue of the parameter estimate information matrix. It is shown how the persistent excitation condition is connected with a proposed concept of stability of a system with uncertain parameters. Unlike similar methods, the proposed algorithms predict the information matrix for more than one step of control. The problem is formulated as an MPC problem with an additional constraint on the information matrix. This constraint makes the problem non-convex, thus only locally optima solutions are guaranteed. The proposed algorithm is derived for ARX system only, but it allows for future reformulation for a general ARMAX system with known moving average (MA) part.

Structured model order reduction of boiler-header models

  • Authors: Sturk, C., Sandberg, H., Trnka, P., prof. Ing. Vladimír Havlena, CSc., Řehoř, J.
  • Publication: Proceedings of the 18th IFAC World Congress, 2011. Bologna: IFAC, 2011. pp. 3341-3347. ISBN 978-3-902661-93-7.
  • Year: 2011
  • DOI: 10.3182/20110828-6-IT-1002.02978
  • Link: https://doi.org/10.3182/20110828-6-IT-1002.02978
  • Department: Department of Control Engineering
  • Annotation:
    This paper presents a model reduction of a boiler-header system. Since it is desirable that the reduced model retains the structure of the full model where the boilers are interconnected with the header, a structured model reduction technique is applied, which takes the entire system into account. This method requires the solution of two linear matrix inequalities to obtain the structured Gramians of the system, but in general it is not possible to guarantee feasibility of these linear matrix inequalities. However for stable systems that are connected in series with a negative feedback-loop with strictly positive real subsystems, we prove that solutions always exist. By showing that the boiler-header system belongs to this class of systems it follows that the structured model reduction method can be applied regardless of the system parameters. © 2011 IFAC.

Grey-box model identification - control relevant approach

  • Authors: Řehoř, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: IFAC Workshops - ALCOSP 2010. Istanbul: Kudret Basim Ambalaj A.S., 2010. pp. 1-4. ISBN 978-3-902661-85-2.
  • Year: 2010
  • Department: Department of Control Engineering
  • Annotation:
    Grey-box modeling is an advantageous tool for system identification when obtained input/output experimental data are insufficiently excited. The lack of information in the data can be often replaced with some additional knowledge about the modeled system, which constricts the class of models under consideration. The real system is usually more complex and do not fit the model class, thus a bias error occurs. The main goal of this paper is to show an effective way how to identify grey-box models, which would be relevant in commissioning predictive control.

Model Predictive Control Relevant Identification in the Grey-box Modeling

  • Authors: Řehoř, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of The 9th International Conference Process Control 2010. Pardubice: Universita Pardubice, 2010, ISBN 978-80-7399-951-3.
  • Year: 2010
  • Department: Department of Control Engineering
  • Annotation:
    Model Predictive Control is very popular in advanced control solutions. Successful implementation is usually highly dependent on the quality of given model. In practical use it is often impracticable or costly to obtain sufficiently excited input-output experimental data. It can be very profitably if designer can use any additional prior information about the process and combined it with the available data. This approach is implicitly included in the grey-box model identification. A new effective way how to identify grey-box models that would be relevant in setting predictive control will be shown. The functionality will be demonstrated using real process data.

Model-Based Predictive Control of a Fluidized Bed

  • Authors: prof. Ing. Vladimír Havlena, CSc., Pachner, D.
  • Publication: Power Plant Applications of Advanced Control Techniques. Wien: Verlag ProcessEng Engineering GmbH, 2010. p. 43-68. ISBN 978-3-902655-11-0.
  • Year: 2010
  • Department: Department of Control Engineering
  • Annotation:
    The chapter presents a novel advanced control strategy for Circulating Fluidized Bed (CFB) boilers. The objective of the advanced control and optimization project was to improve stability of key process variables, effectiveness of limestone use and boiler combustion efficiency under emissions limits. The MPC control strategy is based on state-space model with nonlinear state estimator used as inferential sensor of accumulated char and lime inventory in the bed and linear time varying model parameterized by the bed char inventory. The first-principles based model results from a lumped macroscopic description of the coal particle burning and disintegration in the bed and was validated using experimental data. The solution was validated in SINOPEC Shanghai plant on two Foster-Wheeler CFB boilers burning mixture of coal and coke with nominal steam production 310 t/h.

Noise covariances estimation for Kalman filter tuning

  • Authors: Matisko, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: IFAC Workshops - ALCOSP 2010. Istanbul: Kudret Basim Ambalaj A.S., 2010. pp. 1-6. ISBN 978-3-902661-85-2.
  • Year: 2010
  • DOI: 10.3182/20100826-3-TR-4015.00009
  • Link: https://doi.org/10.3182/20100826-3-TR-4015.00009
  • Department: Department of Control Engineering
  • Annotation:
    Kalman filter tuning is based on process and measurement noise covariances that are parameters of Riccati equation. Based on the Riccati equation solution, Kalman gain is calculated and used for state estimator. Noise covariances are generally not known. The latest methods and their modifications were published in 2005 and later. In many parts of technical science the Bayesian approach can be used for various estimation problems. However, many scientists and researchers a priori consider Bayesian principles to be unpractical because in most cases it is very difficult to work with probabilities or likelihood functions. The probability or likelihood functions cannot be solved analytically for most problems. In this paper, we will discuss the performance of some published methods and compare them with the maximum likelihood approach using numerical methods. Properties of different approaches and qualities of maximum likelihood method will be demonstrated.

Overlapping models merging and interconnection for large-scale model management

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 2010 IEEE International Conference on Control Applications. New York: IEEE Control System Society, 2010. ISSN 1085-1992. ISBN 978-1-4244-5363-4.
  • Year: 2010
  • DOI: 10.1109/CCA.2010.5611036
  • Link: https://doi.org/10.1109/CCA.2010.5611036
  • Department: Department of Control Engineering
  • Annotation:
    The application of advanced control methods to large-scale systems in variable industrial environment requires modeling and identification platform capable of keeping global model with description of its uncertainties, building global model from sub-systems, retrieval of sub-models with mutual consistency, model actualization from new sub-models or new data, etc. This article treats the problem of assembling global model for large-scale system from interconnected and possibly overlapping sub-models, i.e. there can be duplicity in the models. The quality of sub-models can also be different and is taken into account. The article presents two new results: merging of multiple models for the same system by using equivalent data and consistent combination of arbitrary connected models with parametric uncertainty into single model by using statistics of random vectors convolution.

A New Discrete Model of the Rotary Kiln

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc., Pachner, D., Rathouský, J.
  • Publication: Proceedings of the 28th IASTED International Conference on Modelling, Identification and Control. Calgary: Acta Press, 2009. pp. 207-213. ISBN 978-0-88986-781-9.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    The rotary calcination kiln for a production of Titanium dioxide is a very interesting dynamical system. The titanium dioxide is mostly available in a crystaline form known as anatase. The aim of the process is to produce the titanium dioxide with the rutile content around 98% that has the right pigmentary properties in contrast to the anatase. The system of the rotary calcination kiln has several distributed variables (solid, water, gas, vapour, and temperatures of them). These variables have a difficult dependence on each other. The paper presents a distributed parameter discrete time model describing the system behaviour and shows the time responses of the proposed model. It was created the model of the process in Matlab and GUI for the system behaviour analysis.

Comparison of Stochastic Optimal Control Strategies - Monte Carlo Approach

  • Authors: Rathouský, J., prof. Ing. Vladimír Havlena, CSc., Štecha, J.
  • Publication: Proceedings of the 28th IASTED International Conference on Modelling, Identification and Control. Calgary: Acta Press, 2009. pp. 242-247. ISBN 978-0-88986-781-9.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    Stochastic discrete time models are usual tools used for description of uncertain systems. In the paper two optimal control strategies are compared: cautious and certainty equivalent LQ optimal control strategies of ARX model. In all cases the mean value of the criterion is minimized. When all uncertainties in the system are respected and optimal control strategies are used, the probability distribution of the criterion is complicated but its form can be obtained only by Monte Carlo approach. The main goal of the paper is to estimate the form of the criterion distribution.

Grey box model identification with practical application to combustion control

  • Authors: Řehoř, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Identifikace grey-box modelu s praktickou aplikací v průmyslu a energetice. Košice: Technical University of Košice, 2009, ISBN 978-80-553-0237-9.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    The problem of estimating parameters of linear models from noisy time series measurements of continuous dynamical systems is considered. Additional prior information (grey box) is used to improve the quality of the estimate, which can be useful, when obtaining sufficiently excited input/output experiment data is impracticable or costly. The problem is solved with respect to optimal multistep prediction for better performance in advanced control (model predictive control). This leads to a nonconvex numerical optimization, where the solver can easily run out in one of the many local extremes. Two approaches, how to improve the convergence, are introduced. Finally an example of identification for combustion control, based on real power plant data, is presented.

Identification of the gray-box model with practical applications in industry and energy

  • Authors: Řehoř, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 5th International Scientific Symposium on Electric Power Engineering ELEKTROENERGETIKA 2009. Košice: TU Košice, FEI, 2009, pp. 419-423. ISBN 978-80-553-0401-4.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    The problem of estimating parameters of linear models from noisy time series measurements of continuous dynamical systems with multiple inputs and multiple outputs is considered. Additional prior information (grey box) is used to improve quality of the estimate, which can be useful, when obtaining sufficiently excited input/output experimetal data is impracticable or costly. The problem is solved with respect to optimal multistep prediction for better performance in advanced control (model predictive control). This leads to a nonconvex numerical optimization, where the solver can easily run out in one of the many local extremes. Two approaches, how to improve the convergence, are introduced. Finally an example of identification, based on real power plant data, is presented.

Kalman Filter for Systems with Communication Delay

  • Authors: Baramov, L., Pachner, D., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the First IFAC Workshop on Estimation and Control of Networked Systems. Padova: University of Padova, 2009. pp. 310-315. ISBN 978-3-902661-52-4.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    This paper deals with estimating the process state where measurements are obtained via wireless networks. Transmission between the sensor and the estimator may be subject to random delays and/or data losses. A straightforward approach to address this problem is using a time-varying Kalman filter (KF) for the plant augmented by a delay model, which is computationally extensive. Research effort has recently focused on low complexity approximations of these filters. We revisit the time varying KF and, by exploiting the structure of the augmented process model, propose a new algorithm, which is computationally less extensive than the standard one. Moreover, interesting properties of the variable-delay estimators were obtained that are of independent interest.

Numerical Model of Pulverized Coal Combustion in Furnace

  • Authors: Straka, R., Makovička, J., Beneš, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of 7th World Conference on Experimental Heat Transfer, Fluid Mechanics and Thermodynamics. Krakow: AGH-UST, 2009. pp. 717-724. ISBN 978-83-7464-235-4.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    Our main motivation of the combustion model research is its use for design of the combustion chamber geometry and other important parameters needed for the boiler operation.

One-step active control strategy as approximation of dual control

  • Authors: Rathouský, J., prof. Ing. Vladimír Havlena, CSc., Štecha, J.
  • Publication: Preprints of ROCOND'09. Haifa: Technion-israel Institute of Technology, Faculty of EE, 2009. pp. 132-137. ISBN 978-3-902661-45-6.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    Stochastic adaptive control gives a possibility to deal with uncertainties in system descriptions. In contrast to other robust methods, it uses probabilistic description of uncertain system parameters and consequently, stochastic optimization methods are used to design the controller. It also uses identification methods to improve the system model and thus further improve overall performance. In stochastic adaptive control, the controller that achieves required control performance and keeps gathering information about the system at the same time, is referred to as a controller with dual properties. As the optimal dual controller is computationally intractable, approximations of the optimal problem are searched. In this paper we propose a control strategy for ARX systems with dual properties. This active control strategy is based on the well known cautious strategy, but takes the quality of identification in one step ahead into consideration.

ONE-STEP ACTIVE CONTROLLER WITH DUAL PROPERTIES

  • Authors: Rathouský, J., prof. Ing. Vladimír Havlena, CSc., Štecha, J.
  • Publication: 17th International Conference on Process Control´09. Bratislava: Slovak University of Technology, 2009, pp. 453-458. ISBN 978-80-227-3081-5.
  • Year: 2009
  • Department: Department of Control Engineering
  • Annotation:
    In stochastic adaptive control, the controller that achieves required control performance and keeps gathering information about the system at the same time, is referred to as a controller with dual properties. As the optimal dual controller is computationally intractable, approximations of the optimal problem are searched. In this paper we propose a control strategy for ARX systems with dual properties. This active control strategy is based on the well known cautious strategy, but takes the quality of identification in one step ahead into consideration. This strategy shows how to improve control performance mainly in cases when the initial uncertainty in system parameters is large.

Subspace like identification incorporating prior information

  • Department: Department of Control Engineering
  • Annotation:
    The subspace identification methods can further benefit from the prior information incorporation algorithm proposed in this paper. In the industrial environment, there is often some knowledge about the identified system, which can be used to improve the model quality and its compliance with first principles. The proposed algorithm has two stages. The first one is similar to the subspace methods as it uses their interpretation as an optimization problem of finding parameters of an optimal multi-step linear predictor for the experimental data. The second stage with state space model realization from the posterior impulse response estimate is different from the standard subspace methods as it is based on the SWLR (structured weighted lower rank) approximation, which is necessary to preserve the prior information incorporated in the first stage.

An Approach to Out-of-Sequence Measurements in Feedback Control Systems

  • Authors: Pachner, D., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of 2008 IEEE International Workshop on Factory Communication Systems. Piscataway: IEEE, 2008. ISBN 978-1-4244-2349-1.
  • Year: 2008
  • DOI: 10.1109/WFCS.2008.4638704
  • Link: https://doi.org/10.1109/WFCS.2008.4638704
  • Department: Department of Control Engineering
  • Annotation:
    Complex communication network architectures are becoming more frequent in control applications. In such networked control systems, it is often the case that information on process variables is received out-of-time-order. This paper presents a Bayesian approach to handling this out-of-sequence information problem. Such approach leads to a solution involving the joint probability density of current state and past measurements not vet received Under linear Gaussian assumptions the Bayesian solution reduces to an augmented state Kalman Filter. Our approach augments the state dynamically based on the list of missing observations. As this solution can be time and memory, consuming, two simplified implementations of the algorithm are presented.

Multivariable predictive circulating fluidized bed combustor control

  • Department: Department of Control Engineering
  • Annotation:
    Advanced Combustion Control system for the circulating fluidized bed boilers is described. The CFB model is represented by a system of non-linear ordinary differential equations which capture the key CFB behavior. The model exhibits strong cross-interactions of process variables and inverse step response. It is shown the CFB can be successfully controlled using the multiple-inputs multiple-outputs (MIMO) model.

Bayesian Approach to Estimation of Time-Variable Parameters

  • Authors: Tonner, J., Vašíček, O., Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: IFAC Symposium on Computational Economics and Financial & Industrial Systems. Istanbul: Dogus University of Istanbul, 2007,
  • Year: 2007
  • Department: Department of Control Engineering
  • Annotation:
    The aim of this article is to identify structural changes in economy by time variable parameters estimation of Hansen Real Business Cycle model of real economy via modified Extended Bootstrap Filter Smoother. The incorporated rational expectations problem is solved by Generalized Schur Decomposition which is specially adjusted for Bootstrap filter running.

Hierarchical Solution of the Optimum Allocation Problem - Lagrangian Relaxation and Explicit Solution Principles

  • Authors: Cepak, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: In Preprints of the 11th IFAC/IFORS/IMACS/IFIP Symposium on Large Scale Systems (Theory and Applications. Gdansk: Gdansk University of Technology, Faculty of Electrical and Control Engineering, 2007,
  • Year: 2007
  • Department: Department of Control Engineering
  • Annotation:
    Recently, there have been many problems describing the technological, societal and environmental processes, very complex, large in dimension, with many inputs and outputs. These problems can be optimized by classical theories but often such methods based on centrality cannot be applied because of huge amount of information. Then, hierarchical approaches seem to be a convenient tool for the optimization problems. This contribution describes the hierarchical solution of optimum load allocation problem in thermal power plant. Two iterative and one explicit method based on the price coordination and interaction balance are derived and applied.

Integrating Prior Information into Subspace Identification Methods

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of 2007 IEEE Muti-conference on Systems and Control. Singapore: National University of Singapore, 2007. ISSN 1085-1992.
  • Year: 2007
  • Department: Department of Control Engineering
  • Annotation:
    Integrating prior information into subspace identification methods improves their usability for industrial data, where experimental data by them self are in many cases not good enough to give a proper model. The identification experiments in the industrial environment are limited by the economical and safety reasons. However, in practical applications, there is often strong prior information about the identified system, which can be exploited in the identification. The presented algorithm formulates subspace identification as a multi-step predictor optimization. Reformulation to the Bayesian framework allows to incorporate prior information. The paper is completed with the application to the experimental data from the oil burning steam boiler with the rated power of 100 MW.

Model for Estimating Stress in Pressurized Boiler Components

  • Authors: Baramov, L., Beneš, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 46th IEEE Conference on Decision and Control (CDC), 2007. Piscataway: IEEE, 2007. ISSN 0191-2216. ISBN 978-1-4244-1497-0.
  • Year: 2007
  • Department: Department of Control Engineering
  • Annotation:
    This paper proposes a method of modeling stress in pressurized boiler components for the use in a boiler life monitoring system and/or in predictive life-extending control. A low-dimensional model of stress components and temperature is obtained by a suitable approximation of the underlying partial differential equations. A typical boiler component, e.g., a steam header, is spatially large with repeated elements like tubes branching out of the main vessel. The proposed method is based on splitting the component into elementary parts, modeled separately as n-port systems and then obtaining the overall model as an interconnected network. The interconnection is done in the frequency domain to avoid the complexity escalation. Transfer function model is then fitted on the frequency domain data. The resulting model is of low order with a good agreement with a finite element model.

Subspace Identification Incorporating Prior Information

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 46th IEEE Conference on Decision and Control (CDC), 2007. Piscataway: IEEE, 2007, ISSN 0191-2216. ISBN 978-1-4244-1497-0.
  • Year: 2007
  • Department: Department of Control Engineering
  • Annotation:
    Subspace identification methods proved to be a powerful tool, which can further benefit from the incorporation of prior information. In the industrial environment, there is often strong prior information about the identified system, that can be used to improve the model quality and its compliance with physical reality. Such prior information can be the known static gains, the dominant time constants, the impulse response smoothness, etc. An idea comes from the possibility to consider the subspace identification as an optimization problem of finding a model with the optimal multi-step predictions on the experimental data. Further, the problem is reformulated to the Bayesian framework allowing to combine available prior information with the information contained in the experimental data by covariance matrix shaping. The paper is completed with an application to experimental data from an oil firing steam boiler with the rated effective power of 100 MW.

Advanced Control and Real Time Optimization for Power & Industrial Energy

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of International Control Conference. Glasgow: University of Strathclyde, 2006,
  • Year: 2006

Base Vectors for Solving Partial Differential Equations

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc., Beneš, M.
  • Publication: Proceedings of International Control Conference. Glasgow: University of Strathclyde, 2006,
  • Year: 2006
  • Department: Department of Control Engineering
  • Annotation:
    The distributed parameters systems can be described by linear two- dimensional (dependent on two spatial directions) parabolic partial differential equations. Using the finite difference method a distributed parameters system can be transformed to a linear discrete state space model. A controller design based on this description is complicated and for advanced control can be impossible because of the large dimension of the model. Therefore a model reduction method has to be used. We transform the state space model to the balanced realization of the system and show that the state vector of the model can be expressed as the series of columns of the transformation matrix. These columns can be imagined as base vectors of the state space.

Model of Turbulent Coal Combustion

  • Authors: Makovička, J., Beneš, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: COE Lecture Note Series: Kyushu University. 2006, 2005(3), 95-105. ISSN 1881-4042.
  • Year: 2006
  • Department: Department of Control Engineering
  • Annotation:
    We describe behavior of the air-coal mixture using Navier-Stokes equations for gas and particle phases, accompanied by a turbulence model. The undergoing chemical reactions are described by Arrhenian kinetics (reaction rate proportional to exp) ; where T is temperature). We also consider the heat transfer via conduction and radiation. The system of PDEs is discretized using the Finite Volume Method and an Advection Upstream Splitting Method as the Riemann solver. The resulting ODEs are solved using the 4th order Runge-Kutta method. Sample simulation results for typical power production levels are presented.

New Price Coordination Methods Solving Optimum Allocation Problem

  • Department: Department of Control Engineering
  • Annotation:
    Technological processes with many inputs and outputs represent large-scale systems which are often difficult to optimize. It is mainly caused by high dimension of the optimization problem; standard centralized techniques here fail and cannot be employed. The other way how to solve the issue consists in the utilization of the hierarchical approaches. Method of the price coordination constitutes an interesting tool for the solution of the optimum allocation problem. It is based on the interaction balance between the interconnected subsystems. But the problematic of the interaction balance is defined only on the theoretical field and some practical implementations are not discussed in the literature. In this contribution, there are three newly designed algorithms based on the theoretical aspects of the interaction balance theorem. Next, they are applied on the optimum allocation problem and the results are compared to the ones obtained by standard centralized techniques.

Range Control MPC Approach for Two-Dimensional System

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: CTU Reports - Proceedings of Workshop 2006. Praha: Česká technika - nakladatelství ČVUT, 2006. pp. ?. ISBN 80-01-03439-9.
  • Year: 2006
  • Department: Department of Control Engineering
  • Annotation:
    This paper deals with two-dimensional dynamic processes (systems with parameters dependent on two spatial directions) which can be described by lumped inputs and distributed output models. These models can be mathematically described by partial differential equations [1]. Unlike ordinary differential equations, the partial differential equations contain, in addition, derivatives with respect to spatial directions. Consequently, the partial differential equations lead to more accurate models but their complexity is larger.

Recursive Subspace Identification Algorithm

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Radio-Electronics, Electrical and Power Engineering. Moscow: Moskovskij energeticeskij institut, 2006, pp. 483-484. ISBN 5-87789-022-0.
  • Year: 2006
  • Department: Department of Control Engineering
  • Annotation:
    Modern control methods, like model predictive control, proved to be very effective in the industrial applications. However, the efficiency is often limited by the quality of the controlled system model, which is hard to obtain especially for the large systems with multiple inputs and multiple outputs.The recent advances in Subspace identification methods (4SID) showed that they could be successful in the identification of such models from real world data. We propose a new recursive 4SID method in the well-known least squares framework. The 4SID methods can be shown to give a model, which is an optimal multi-step predictor, in the sense of minimizing the sum of prediction errors on the measured input/output data for a certain prediction horizon. This fact can be exploited for recursive 4SID algorithm, allowing us to use recursive least squares with some type pf forgetting and even allows us to incorporate prior information in the field of otherwise black-box approach of 4SID methods.

Recursive Subspace Identification in the Least Squares Framework

  • Department: Department of Control Engineering
  • Annotation:
    Subspace identification methods (4SID) proved to be efficient for industrial applications, due to their good properties, such as: same complexity of identification for single input/output and multiple input/output systems, direct state space model identification, numerical robustness (QR and SVD factorization) and implicit model order reduction. The algorithms are well developed for off-line identification, however on-line recursive identification is still rather an open topic. The problem lies in the recursification of SVD, which is impossible and several approximations are used instead. We use a different approach, exploiting the fact, that 4SID methods minimizes implicit optimality criterion, which is mean square error of multi-step predictions of the model. The criterion allows for recursification in the least squares framework and prior knowledge incorporation. We also address the problem of non-causality, which was recently pointed out in 4SID methods.

Robust Range Control for LTI Systems with Bounded Disturbance Input

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of International Control Conference. Glasgow: University of Strathclyde, 2006.
  • Year: 2006
  • Department: Department of Control Engineering
  • Annotation:
    The paper presents an algorithm for control of linear time-invariant dynamical systems with bounded disturbance input. The proposed algorithm is based on range control which belongs to the family of model based predictive controllers. Range control can be regarded as a generalized form of a tracking problem where the set-point for the system output is replaced by a funnel (time-varying lower and upper limits on the system output in the form of soft constraints).

The Development of the Steam Condenser Model for the Simulation of the Condensing Turbine Operation

  • Department: Department of Control Engineering
  • Annotation:
    The steam condensing turbines together with the backpressure units represent one of the basic devices used for the generation of the electrical energy. The main difference between them consists in the fact that the condensing turbines contrary to the backpressure ones operate at theirs outputs with the pressures located very close to the vacuum. This is the reason why they are applied as the last equipments in the multipart steam turbines. The indivisible part of the condensing turbine is constituted by the steam condenser where the flowing water cools the steam passing through turbines; by this way it liquefies. While the modeling of the condensing turbine does not differ too much from the backpressure one, the condensation process seems to be quite complicated event. This paper describes the development of the mathematical model of the condensing unit for the simulation of the condensing turbines operation in MATLAB. The next purpose is the testing of the designed model in the SIMULI

A Brief Introduction To Control Design Demonstrated On Laboratory Model Servo DR300 - AMIRA

  • Department: Department of Control Engineering
  • Annotation:
    This paper is written as a motivation for the students who are beginning to study the control engineering. Therefore, the below mentioned references are introduced with regard to this fact. The paper shortly presents the procedure of the control design, including a description of a system, an identification of its parameters, a simple and an advanced control-ler design. The controller design procedures are illustrated on the laboratory model servomechanism DR300 - AMIRA, which is held in the laboratory of control theory K26, Department of Control Engineering, Faculty of Electrical Engineering, Czech Technical University in Prague

A Distributed Automation Framework for Plant-Wide Control, Optimisation, Scheduling and Planning

  • Authors: prof. Ing. Vladimír Havlena, CSc., Lu, J.
  • Publication: Proceedings of the 16th World Congress of the International Federation of Automatic Control. Praha: IFAC, 2005. ISSN 1474-6670. ISBN 978-0-08-045108-4.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    The objective of the talk will be to identify current open problems and trends in plant wide control and demonstrate a solution based on distributed, solution component based architecture for integrated process management, embracing the layers of Advanced Process Control, Real Time Optimisation and Planning & Scheduling, in selected application areas. The problems and outlined solutions are intended to stimulate discussion as well as attract more research interest.

Application of Model Predictive Control to Advanced Combustion Control

  • Department: Department of Control Engineering
  • Annotation:
    The objective of the application of model-based predictive control technology for boiler control is to enable tight dynamical coordination of selected controlled variables, particularly the coordination of air and fuel flows during transients. It is shown that this approach can be used in connection with excess air optimization to increase boiler efficiency while considerably reducing the production of NOx

Base Vectors for Solving Partial Differential Equations

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Technical Computing Prague 2005: 13th Annual Conference Proceedings. Praha: VŠCHT, 2005, ISBN 80-7080-577-3.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    The distributed parameters systems can be described by linear two-dimensional (dependent on two spatial directions) parabolic partial differential equations. Using the finite difference method a distributed parameters system can be transformed to a linear discrete state space model. The controller design based on this description is complicated because of the large dimension of the model. Therefore, a model reduction method has to be used. We transform the state space model to the balanced realization of the system and show that the state vector of the model can be expressed as the series of columns of the transformation matrix. These columns can be imaged as base vectors of the state space.

Bootstrap Filtering for Czech Macro-economic Model Estimation

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc., Štecha, J.
  • Publication: 15th International Conference on Process Control 05. Bratislava: Slovak University of Technology, 2005, ISBN 80-227-2235-9.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    Implementing direct Bayesian inference using Monte Carlo methods (Bootstrap filter) we identified Czech macro-economic model based on the work (Clarida et al. 1999). The main concern was to identify model parameters for the prediction of model behavior, which is essential for taking proper economical decisions. Simultaneous estimation of model parameters led to non-linear model. Commonly used Extended Kalman filter failed in this case, therefore we used bootstrap filter, which can handle non-linear and/or non-gaussian systems. The posterior probability density function of states and parameters were obtained from the prior probabilities (represented as a large set of samples), which were updated from measured data according to Bayesian inference. Given only limited data set (quarterly data from 1994) at disposal we incorporated smoothing (backward filtration) into bootstrap filter to maximize the use of information from the data.

Control Design for Servo AMIRA-DR300

  • Authors: Roubal, J., Augusta, P., prof. Ing. Vladimír Havlena, CSc., Fuka, J.
  • Publication: 15th International Conference on Process Control 05. Bratislava: Slovak University of Technology, 2005, ISBN 80-227-2235-9.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    This paper is written for students who are starting to study at Department of Control Engineering, CTU in Prague as the motivation for studying control engineering. The paper shortly presents procedures of the control design, includinga description of a system, an identification of its parameters, a simple and an advanced controller design. The controller design procedures are illustrated in the control design for the laboratory model servomechanism AMIRA -- DR300.

Control of Distributed Paramerets System by MPC Controller

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of Workshop 2005. Praha: České vysoké učení technické v Praze, 2005, pp. 172-173. ISBN 80-01-03201-9.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    This paper deals with model predictive control of a distributed parameters system which is described by a linear two-dimensional (dependent on two spatial directions) parabolic partial differential equation. This partial differential equation is transformed to the discrete state space description using the finite difference approximation. A model with a large dimension is obtained and has to be reduced for an advanced control design. The balanced truncation method is used for the model dimension reduction. For this low dimension model, the range control approach is applied.

Control of Distributed Paramerets System by MPC Controller

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of Workshop 2005 - Part A,B. Praha: České vysoké učení technické v Praze, 2005, pp. 172-173. ISBN 80-01-03201-9.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    This paper deals with model predictive control of a distributed parameters system which is described by a linear two-dimensional (dependent on two spatial directions) parabolic partial differential equation. This partial differential equation is transformed to the discrete state space description using the finite difference approximation. A model with a large dimension is obtained and has to be reduced for an advanced control design. The balanced truncation method is used for the model dimension reduction. For this low dimension model, the range control approach is applied.

Control of System Represented by Multiple Model

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc., Roubal, J.
  • Publication: Journal of Electrical Engineering. 2005, 56(11-12), 281-289. ISSN 1335-3632.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    The optimal control strategy for a continuous and discrete time multiple model is developed. Simulation of control and on-line estimation of model probability is shown. Robustness of stability and comparison of the classical control of a Linear state space model with Quadratic criterion (LQ control) and LQ control based on multiple models are presented.

Design and Analysis of Model based Predictive Controllers

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of Workshop 2005. Praha: České vysoké učení technické v Praze, 2005, pp. 174-175. ISBN 80-01-03201-9.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    In this contribution we present application of model-based predictive controller based on mixture distribution, i.e. a set of parallel models, to CSTR process. This process is highly non-linear. Presented MPC is based on a mixture of linearized models in several operating points parameterized by their probabilities estimated on-line. Provided that the operating point of the process lies within the given set defined by convex envelope of the operating points presented MPC ensures "soft switching" between the individual models. The on-line estimation of model probabilities is done by a set of Kalman filters in the normalized form. The good performance of algorithm for model probability estimation is very important part of presented MPC algorithm because the resulting MPC is parameterized by them.

Development of the Steam Condenser for the Simulation of the Condensing Turbine Work

  • Authors: Cepák, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Technical Computing Prague 2005: 13th Annual Conference Proceedings. Praha: VŠCHT, 2005. ISBN 80-7080-577-3.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    This contribution describes the modeling of the steam condenser and its following application for the simulation of the condensing turbine work. Developed model is validated by the real data and then implemented into the library of the energetic components containing early designed units as turbine part, injection cooler and steam header. Convenient combination of these construction elements leads to the faithful simulation of the whole steam turbines; afterwards, it is very easy to observe their work.

Enhancing ARX-Model Based MPC by Kalman Filter and Smoother

  • Authors: Baramov, L., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 16th World Congress of the International Federation of Automatic Control. Praha: IFAC, 2005. ISSN 1474-6670. ISBN 978-0-08-045108-4.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    An approach to enhancing a model-based predictive controller by Kalman filter is proposed. The controller uses an ARX process model and the structure of the controller is assumed fixed; some of its internal variables - past values of controlled variables (output history) are accessible and can be modified to achieve better performance in disturbance attenuation and noise rejection. We present an algorithm of updating the output history using Kalman filter to achieve predictions equivalent to those of the statespace model, thus overcoming the limitations of the ARX predictor. Interesting relations of this algorithm to Kalman interval smoother are given.

Estimation of the Czech Macroeconomic Model by Bootstrap Filter

  • Authors: Trinkewitz, J., Štecha, J., prof. Ing. Vladimír Havlena, CSc., Vašíček, O., Pytelková, H.
  • Publication: Proceedings of the 16th IASTED International Conference on Modelling and Simulation. Calgary: Acta Press, 2005. pp. 203-208. ISBN 0-88986-498-5.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    The monetary policy problem is explained in a simple theoretical framework of economy. The article presents a theoretical macroeconomic model and it shows the theoretical procedure of finding the optimal monetary policy under discretion. The model is quantitatively analyzed on the data of the Czech economy. Model parameters are estimated simultaneously by the original "extended bootstrap filter smoother".

Finite Volume Numerical Model of Coal Combustion

  • Authors: Makovička, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of Czech - Japanese Seminar in Applied Mathematics 2004. Praha: České vysoké učení technické v Praze, Fakulta jaderná a fyzikálně inženýrská, 2005, pp. 106-116. ISBN 80-01-03181-0.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    We describe behavior of the air-coal mixture using two sets of Navier-Stokes equations for gas and particle phases. The undergoing chemical reactions are described by Arrhenian kinetics. We also consider the aerodynamic forces between the gas and particle phases, and the heat transfer via conduction and radiation. The system of PDEs is discretized using the Finite Volume Method and an Advection Upstream Splitting Method as the Riemann solver. The resulting ODEs are solved using the 4th order Runge-Kutta method. Several simulation results usually leading to stationary states for different energy production levels of the furnace are presented.

Minimum Order Transfer Function: the Interpolation Approach

  • Department: Department of Control Engineering
  • Annotation:
    This paper presents an algorithm for obtaining the minimum order MISO transfer function model for the use in a model-based predictive controller. The source model can be either a non-minimal ARX model, a state-space model or any interconnection of linear models of mixed state-space and transfer function representations. The algorithm is based on polynomial interpolation theory, representing polynomials by their values on a set of points in the complex plane. Using this theory, we can find the minimum order from the dimension of the null space of a particular matrix. Finding the minimum order model is equivalent to finding a specific base of the null space. A novel feature of the presented approach is using a set of complex interpolation nodes obtained by mapping the standard set of real Chebyshev nodes by a bilinear transform.

Model Predictive Control with Invariant Sets: Set-Point Tracking

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 15th International Conference on Process Control 05. Bratislava: Slovak University of Technology, 2005, ISBN 80-227-2235-9.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    The stability of a control loop with the model based predictive controller (MPC) is recently closely connected to concept of invariant sets. In many control applications, the system output should track a given reference signal. In this paper, the extension of the basic MPC algorithm to reference tracking is shown. Main result of the paper is that the minimal and maximal allowed bounds of the reference signal, for which the optimization problem of reference tracking remain feasible, are piecewise affine functions of the system state. For each length of the prediction horizon, the proposed problem can be solved as a pair of multi-parametric linearprograms.

On a Model of Coal Combustion

  • Authors: Beneš, M., prof. Ing. Vladimír Havlena, CSc., Makovička, J.
  • Publication: Algoritmy 2005 - Proceedings of contributed papers and posters. Bratislava: Slovak University of Technology, 2005. pp. 12-21. ISBN 978-80-227-2192-9.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    We describe behavior of the air-coal mixture using two sets of Navier-Stokes equations for gas and particle phases. The undergoing chemical reactions are described by Arrhenian kinetics We also consider the aerodynamic forces between the gas and particle phases, and the heat transfer via conduction and radiation. The system of PDEs is discretized using the Finite Volume Method and an Advection Upstream Splitting Method as the Riemann solver. The resulting ODEs are solved using the 4th order Runge-Kutta method. Results regarding numerical convergence estimation and parallelization effciency are presented.

Platform for Advanced Control Applications

  • Authors: prof. Ing. Vladimír Havlena, CSc., Findejs, J., Beran, J., Horn, B., Rozložník, M.
  • Publication: Proceedings of the 16th World Congress of the International Federation of Automatic Control. Praha: IFAC, 2005. ISSN 1474-6670. ISBN 978-0-08-045108-4.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    Control applications have many requirements not provided by commercial operating systems. This paper describes the characteristics and usage of an environment for hosting process-control applications, which is implemented on a commercial operating system. It is called Unified Real Time (URT) platform, and is intended for applications that are large or complex and that may involve dynamic configuration, flexible scheduling, complex organization, etc. This paper also demonstrates the structure of a typical Advanced Control Application (ACA) designed under URT.

Sensitivity Analysis and Robust Model Predictive Control of Singular Systems

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 7th IASTED International Conference on Control and Applications. Calgary: Acta Press, 2005. pp. 95-100. ISBN 0-88986-502-7.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    Numerous industrial processes can be described by a set of nonlinear differential algebraic equations. In the control engineering, such systems are known as singular systems. In this paper, we propose a robust model predictive control scheme for nonlinear singular systems based on sensitivity analysis. The sensitivity analysis rests in computing the sensitivity matrices for the system output along the prediction trajectory. The proposed approach leads to the framework known as Sequential Quadratic Programming.

Subspace Identification as Multi-step Predictions Optimization

  • Authors: Trnka, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the Fifth IASTED International Conference on MODELLING, SIMULATION, AND OPTIMIZATION. Anaheim: ACTA Press, 2005. pp. 223-228. ISBN 0-88986-524-8.
  • Year: 2005
  • Department: Department of Control Engineering
  • Annotation:
    We will show that apparently complicated and not easy to understand expressions with geometrical projections appearing in the algorithms of subspace identification for linear state space models, can be quite simply derived from optimizing these models for multi-step predictions on measured data samples using least squares. Furthermore we will show the advantages, which bring the use of multi-step predictions instead of single-step predictions arising from regression methods.

Volume Numerical Model of Coal Combustion

  • Authors: Makovička, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of Czech - Japanese Seminar in Applied Mathematics 2004. Praha: České vysoké učení technické v Praze, Fakulta jaderná a fyzikálně inženýrská, 2005. pp. 106-116. ISBN 80-01-03181-0.
  • Year: 2005

Application of nonlinear data reconciliation method for steam generators data

  • Authors: Vítek, T., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Ingineering mechanics 2004, National conference with international participation. Praha: Ústav termomechaniky AV ČR, 2004.
  • Year: 2004

Control of CSTR Using Model Predictive Controller Based on Mixture Distribution

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 6th IFAC-Symposium on Nonlinear Control Systems. Düsseldorf: VDI/VDE Mess- und Automatisierungstechnik, 2004. p. ?.
  • Year: 2004

Design and Analysis of Model Predictive Control using MPT Toolbox

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: MATLAB 2004 - Sborník příspěvků 12. ročníku konference. Praha: VŠCHT, 2004, pp. 468-476. ISBN 80-7080-550-1.
  • Year: 2004
  • Department: Department of Control Engineering
  • Annotation:
    The Multi-Parametric Toolbox (MPT) is a useful toolbox for Matlab that helps by computing explicit optimal control law for constrained systems. The recent concept of Model Predictive Control (MPC) that ensures closed-loop stability uses the polytopic or ellipsoidal invariant sets. The MPT toolbox provides a number of algorithms for computing with polytopic sets. In this paper, the MPT toolbox is used for designing of MPC controller that guarantees closed-loop stability and asymptotic reference tracking.

Identification and Predictive Control of ARX Model by p Norm Minimization

  • Authors: Pekař, J., Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 23rd IASTED International conf. on Modelling Identification, and Control. Zürich: Acta Press, 2004. p. 62-67. ISBN 0-88986-387-3.
  • Year: 2004

Invariant Sets in Model Predictive Control

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 8th International Student Conference on Electrical Engineering, POSTER 2004, May 20 2004, Prague. Praha: ČVUT v Praze, FEL, 2004,
  • Year: 2004

Library of Energetic Components for Simulation of Steam Turbines

  • Authors: Cepák, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: MATLAB 2004 - Sborník příspěvků 12. ročníku konference. Praha: VŠCHT, 2004, pp. 60-63. ISBN 80-7080-550-1.
  • Year: 2004

Mathematical model of the backpressure multistage steam turbine

Modelling of the multipart steam turbine in Matlab environment

  • Authors: Cepák, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 15th International Conference on Systems Science. Wroclaw: Oficyna Wydawnicza Politechniki Wroclawskiej, 2004. pp. 410-417. ISBN 83-7085-804-X.
  • Year: 2004

MPC Controller for Two-Dimensional Systems

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc., Neuhauser, J., Trnka, P.
  • Publication: Proceedings of the 15th International Conference on Systems Science. Wroclaw: Oficyna Wydawnicza Politechniki Wroclawskiej, 2004. pp. 132-140. ISBN 83-7085-804-X.
  • Year: 2004
  • Department: Department of Control Engineering
  • Annotation:
    This paper deals with the model predictive control of a distributed parameter system which is described by a linear two-dimensional (dependent on two spatial directions) parabolic partial differential equation. This partial differential equation is transformed to the discrete model using the finite difference approximation. The discrete model with large dimension is transformed using the subspace identification method to the state space model with low dimension. This model is used for the model predictive control approach.

Predictive Control of Distributed Parameter System

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: 8th International Student Conference on Electrical Engineering, POSTER 2004, May 20 2004, Prague. Praha: ČVUT v Praze, FEL, 2004,
  • Year: 2004
  • Department: Department of Control Engineering
  • Annotation:
    The predictive control strategy based on the lumped input/output model is used as the standard approach for the slow industrial processes. This paper deals with the MPC controller design for the linear one-dimensional distributed parameter system which is described by the parabolic partial differential equation. At first, the partial differential equation is converted to the discrete description by the finite element method and then the standard linear state space description is developed.

Predictive Controller for Mixture of Output Error Models

  • Authors: prof. Ing. Vladimír Havlena, CSc., Pekař, J., Štecha, J.
  • Publication: Proceedings of the 23rd IASTED International conf. on Modelling Identification, and Control. Zürich: Acta Press, 2004. p. 438-443. ISBN 0-88986-387-3.
  • Year: 2004

Predictive Model of Two-Dimensional System

  • Authors: Roubal, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 6th International Scientific-Technical Conference on Process COntrol (Říp 2004). Pardubice: Univerzita Pardubice, 2004, pp. 138. ISBN 80-7194-662-1.
  • Year: 2004
  • Department: Department of Control Engineering
  • Annotation:
    This paper deals with the predictive model of a distributed parameter system which is described by a linear two-dimensional (dependent on two spatial directions) parabolic partial differential equation. This partial differential equation is transformed to the discrete state space description using by the finite difference approximation. This model is used for the prediction of the system behaviour.

Simulation of the Energetic Components in Matlab Environment

  • Authors: Cepák, M., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 6th International Scientific-Technical Conference on Process COntrol (Říp 2004). Pardubice: Univerzita Pardubice, 2004. pp. 121. ISBN 80-7194-662-1.
  • Year: 2004

Simulation of Two-Dimensional Distributed Parameters Systems

  • Authors: Roubal, J., Trnka, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: MATLAB 2004 - Sborník příspěvků 12. ročníku konference. Praha: VŠCHT, 2004, pp. 500-504. ISBN 80-7080-550-1.
  • Year: 2004
  • Department: Department of Control Engineering
  • Annotation:
    The distributed parameters systems can be described by a linear two-dimensional (dependent on two spatial directions) parabolic partial differential equation. This paper deals with a transformation of this model to the classical linear dynamical state space model which can be used for control design. For accurate description, this model has a large dimension which can produce problems with advanced controller design, for example,with model predictive control approach. Therefore a model reduction has to be used for a controller design. The influence of the model reduction on the model accuracy is discussed.

The Efficient Robust MPC Algorithm: Simulation Results

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the 6th International Scientific-Technical Conference on Process COntrol (Říp 2004). Pardubice: Univerzita Pardubice, 2004, pp. 74. ISBN 80-7194-662-1.
  • Year: 2004

Application of Multiple Model LQ Control

  • Department: Department of Control Engineering
  • Annotation:
    The optimal control strategy for discrete time multiple model was described. Simulation of control and on-line estimation of model probability is shown in this paper. Comparison of the classical LQ control and LQ control based on multiple model is presented.

Design of Model based Predictive Controllers by lp-norm Minimization in MATLAB

Designe of Decentralized Control Using the Dynamic Compensation Method

Influence of Noise Model on k-step ahead Optimal Predictor

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of Process Control '03. Bratislava: Slovak University of Technology, 2003, pp. 39-1-39-6. ISBN 80-227-1902-1.
  • Year: 2003

Model Predictive Controller for Mixtue of State Space Models

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Principia Cybernetica 03. Liberec: Technická univerzita, 2003. p. 93-98. ISBN 80-7083-733-0.
  • Year: 2003

Optimal Control of Multiple Model

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: The 22nd IASTED International Conference on Modelling Identification and Control. Calgary: IASTED, 2003. p. 1-6. ISSN 1025-8973. ISBN 0-88986-339-3.
  • Year: 2003

Robustness of Multiple Model Control

  • Department: Department of Control Engineering
  • Annotation:
    The optimal control strategy for discrete time multiple model is described. Simulation of control and on-line estimation of model probability is shown. Robustness of stability and comparison of the classical LQ control and LQ control based on multiple models is presented.

Combustion Optimization with Inferential Sensing

  • Authors: prof. Ing. Vladimír Havlena, CSc., Findejs, J., Pachner, D.
  • Publication: Proceedings of the 2002 American Control Conference. New York: IEEE, 2002. p. 3890-3895. vol. 1-6. ISSN 0743-1619. ISBN 0-7803-7298-0.
  • Year: 2002
  • Department: Department of Control Engineering
  • Annotation:
    The paper presents development and implementation of Advanced Combustion Controller(ACC) for a coalfired boiler. The solution consists of Combustion Controller (multi-variable predictive controller) and Combustion Optimizer (cautious strategy stochastic optimizer). Optimization id based on a model of the CO and NOx emissions. The model is used to calculate the setpoint of the optimal air/fuel ratio(s) maximizing the effciency of the plant under constraints given by emission limits.

Control and Estimation of parallel Models

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Kybernetika - teória, vyučovanie a prax. Žilina: Žilinská univerzita, 2002, pp. 57-62. ISBN 80-967609-8-X.
  • Year: 2002

Economic Load Allocation

  • Authors: Somvarský, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Preprints of the 15th Triennial World Congress of the International Federation of Automatic Control. Oxford: Elsevier, 2002, pp. 2413-2418.
  • Year: 2002

Lq-Iqg Controller For Parallel Models

  • Authors: prof. Ing. Vladimír Havlena, CSc., Štecha, J.
  • Publication: Preprints of the 15th Triennial World Congress of the International Federation of Automatic Control. Oxford: Elsevier, 2002, pp. 2118-2122.
  • Year: 2002

Mathematical modelling of steam and flue gas flow in a heat exchanger of a steam boiler

  • Authors: Beneš, M., prof. Ing. Vladimír Havlena, CSc., Makovička, J.
  • Publication: Algoritmy 2002. Bratislava: Slovak University of Technology, Faculty of Civil Engineering, 2002. pp. 171-178. ISBN 80-227-1750-9.
  • Year: 2002

Multiple Models Estimation and Control

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: The 6th World Multiconference on Systemics, Cybernetics and Informatics. Orlando: IIIS - International Institute of Informatics and Systemics, 2002. pp. 248-253. ISBN 980-07-8150-1.
  • Year: 2002

Predictive Controller for Mixture of Output Error Models

  • Authors: Pekař, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of Workshop 2002. Praha: České vysoké učení technické v Praze, 2002, pp. 330-331. ISBN 80-01-02511-X.
  • Year: 2002

Predictive Optimal Control Strategy for Parallel Models

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the IASTED International Conference Modelling, Identification, and Control. Zürich: Acta Press, 2002, pp. 240-245. ISSN 1025-8973. ISBN 0-88986-319-9.
  • Year: 2002

Alternative Models in Fault Detection-Monte Carlo Approach

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc., Trinkewitz, J.
  • Publication: Proceedings of the IASTED Inter. Conference MIC. Calgary: IASTED, 2001, pp. 154-158. ISBN 0-88986-316-4.
  • Year: 2001

State Estimation with Bounded Deterministic Errors

  • Authors: Pachner, D., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Poster 2001. Praha: České vysoké učení technické v Praze, Fakulta elektrotechnická, 2001. pp. IC33.
  • Year: 2001

Application of MPC to Advanced Combustion Control

  • Authors: prof. Ing. Vladimír Havlena, CSc., Findejs, J.
  • Publication: IFAC Symposium on Power Plants & Power Systems Control 2000 - Preprints. Brussels: IBRA-BIRA Federation, 2000, pp. 167-173.
  • Year: 2000

Model Predictive Control - Review and Case Study

Numerical Algorithms for Structured PVC Production

Optimization and Control of Combustion Process Products

  • Authors: Pachner, D., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: IFAC Symposium on Power Plants & Power Systems Control 2000 - Preprints. Brussels: IBRA-BIRA Federation, 2000. pp. 174-181.
  • Year: 2000

Advenced Combusiton Control for Pulverized-Coal Fired Boilers

  • Authors: prof. Ing. Vladimír Havlena, CSc., Findejs, J., Jech, J.
  • Publication: PRAGOREGULA 99. Praha: Masarykova akademie, 1999, pp. 35-40. ISBN 80-902131-3-8.
  • Year: 1999

Continous Time Control for PVC Production

  • Authors: Barva, P., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Preprints vol. N, 14th WC of IFAC. London: Elsevier Science, 1999, pp. 115-121. ISBN 0-08-043225-5.
  • Year: 1999

Development of ACC Controller with Matlab/Simulink

  • Authors: prof. Ing. Vladimír Havlena, CSc.,
  • Publication: MATLAB '99. Praha: VŠCHT - Ústav fyziky a měřicí techniky, 1999, pp. 52-59. ISBN 80-7080-354-1.
  • Year: 1999

Identification for Control-Monte Carlo Approach

  • Authors: Štecha, J., Pračka, T., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the IASTED International Conference. Calgary: IASTED, 1999, pp. 101-106.
  • Year: 1999

Nonlinear MPC and Inferential Sensing for PVC Production

  • Authors: prof. Ing. Vladimír Havlena, CSc., Barva, P.
  • Publication: Proceedings of the 1999 CCA/CACSD. Norfolk, VA: Omnipress, 1999, pp. 915-920. ISBN 0-7803-5449-4.
  • Year: 1999

Systems Classification by Bootstrap Filter

  • Authors: Štecha, J., Pračka, T., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Conference Proceedings of European Control Conference ECC99 (on CD). Aachen: Rubicon - Agentur für digitale Medien, 1999, pp. 5-10.
  • Year: 1999

Advanced Control for PVC Production

  • Authors: Barva, P., prof. Ing. Vladimír Havlena, CSc., Horák, J.
  • Publication: Dykops - 5 (IFAC Symposium). Thessaloniki: Aristotle University of Thessaloniki, 1998. pp. 262-267.
  • Year: 1998

Review of State Estimation Methods with Multiple Process Models

Smoothing by Weightend Bootstrap Method

  • Authors: prof. Ing. Vladimír Havlena, CSc., Štecha, J.
  • Publication: The 3rd IEEE European Workshop on Computer Intensive Methods in Control and Data Processing. Praha: AV ČR, 1998, pp. 25-28.
  • Year: 1998

State Estimation of Baker's Yeast Fed Bath Cultivation by Extended Kalman Filter Using Alternative Models

  • Authors: prof. Ing. Vladimír Havlena, CSc., Hrčiřík, P., Náhlík, J.
  • Publication: Dykops - 5 (IFAC Symposium). Thessaloniki: Aristotle University of Thessaloniki, 1998, pp. 632-637.
  • Year: 1998

Receding-Horizon MIMO LQ Controller Design with Guaranteed Stability

Smoothing in Simultaneous State and Parameters Estimation

Tree Structures for Identification with Alternative Models

Decision and Aproximation Based Algorithms for Identification with Alternetive Models

  • Authors: prof. Ing. Vladimír Havlena, CSc., Štecha, J., Kraus, F.
  • Publication: Proceedings of 13th IFAC World Congress. San Francisko: IFAC, 1996, pp. 274-278. ISBN 0-08-042605-0.
  • Year: 1996

Smoothing with Alternative Models of Parameter Development

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc., Vašíček, O.
  • Publication: Modelling Identification and Control. Calgary: IASTED, 1996, pp. 179-182. ISBN 0-88986-193-5.
  • Year: 1996

Internal Model Principle and Asymptotic Reference Signal Tracking

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Third IASTED International Conference on Computer Applications in Industry. Calgary: IASTED, 1995, pp. 107-111. ISBN 0-88986-210-9.
  • Year: 1995

Smoothing in Simultaneous State and Parameter Estimation in a Linear System

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Proceedings of the Third European Control Conference ECC 95. Berlin: Springer, 1995, pp. 2165-2170. ISBN 3-540-19967-5.
  • Year: 1995

Bayerian Order Reduction for ARMAX Systems

  • Authors: prof. Ing. Vladimír Havlena, CSc., Kunc, D.
  • Publication: SYSID 94 - 10th IFAC Symposium on System Identification. Copenhagen: IFAC, 1994, pp. 251-256. ISBN 87-7748-034-1.
  • Year: 1994

Parameter Tracking with Alternative Noise Models

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Computer-Intensive Methods in Control and Signal Processing. Berlin: Springer, 1994, pp. 170-175.
  • Year: 1994

Parametr Identification with Alternative Models

  • Authors: Štecha, J., prof. Ing. Vladimír Havlena, CSc.,
  • Publication: Sborník - setkání kateder. Praha: ČVUT FEL, Katedra řídicí techniky, 1994, pp. 30-35.
  • Year: 1994

Fault Detection Based on Alternative Models of Parameter Development

  • Authors: prof. Ing. Vladimír Havlena, CSc., Štecha, J.
  • Publication: Proceedings of the Int. Workshop on Applied Automatic Control 93. Praha: České vysoké učení technické v Praze, 1993, pp. 44-47.
  • Year: 1993

SIMULTANEOUS PARAMETER TRACKING AND STATE ESTIMATION IN A LINEAR-SYSTEM

  • Department: Department of Control Engineering
  • Annotation:
    The task of simultaneous tracking of time-varying parameters and estimation of the state is treated for a linear system described by a time-varying input-output ARMAX or Delta model with known c (noise) parameters. First, a Bayesian approach-based conceptual solution is presented. Then it is shown that utilizing the properties of the observer canonical state model, algebraic recursion operating on the joint parameter and state mean and covariance matrix can be obtained with no approximation involved. Several illustrative examples are included.

Smoothing Preserving Discontinuity Based on Alternative Models of Parameter Development

  • Authors: prof. Ing. Vladimír Havlena, CSc., Štecha, J., Pajdla, T.
  • Publication: Proceednigs of the Czech Pattern Recognition Workshop. Prague: Czechoslovak Pattern Recognition Society, 1993, pp. 75-83.
  • Year: 1993

Smootling with Alternative Model of Parameter Development

Responsible person Ing. Mgr. Radovan Suk