|
|
Performance, robustness and implementation of controllers
Buchta, Luděk ; Michal, Mrázek (referee) ; Pivoňka, Petr (advisor)
This master‘s thesis deals with the design of modern process control algorithms suitable for SISO industrial control equipments. In this thesis the analysis of the controller design methodology using Mixed-sensitivity function is described. This method consists in shaping frequency characteristics of the sensitivity function and complementary sensitivity fiction using the weighting functions. The proposed H controllers are compared with the classical structure of a discrete PID controller with filtered derivative part. The proposed controllers are compared in terms of robustness, performance, complexity of the design and requirements necessary for their practical application (anti-windup, smooth switchover). Robustness of the controller is evaluated on the basis of modulus stability margin. Estimation of model´s parameters is solved least-square method. The proposed control system consists of an industrial PC and Automation Panel from the company B&R, decentralized system I/O and the real plant. Visualization for control workplace was created in the program Automation studio. This visualization is used to the easy transfer of information from a controlled process.
|
|
|
Adaptive controllers with principles of artificial intelligence and its comparison with classical identifications methods
Dokoupil, Jakub ; Malounek, Petr (referee) ; Pivoňka, Petr (advisor)
This piece of work deals with a philosophy of design adaptive controller, which is based on knowledge of mathematical model controlled plant. This master thesis is focused on closed-loop on-line parametric identification methods. An estimation of model´s parametres is solved by two main concepts: recursive leastsquare algorithms and neural estimators. In case of least-squares algorithm the strategy of preventing the typical problems are solved here. For instance numerical stability, accurecy and restricted forgetting. Back Propagation and Marquardt- Levenberg algorithm were choosen to represent artificial inteligence. There is still a little supermacy on the side of methods based on least-squares algorithm. To compare individual algorithms the grafical interface in MATLAB/Simulink was created.
|
|
|
Quadratically Optimal Augmented Identification and Filtration
Dokoupil, Jakub ; Bobál, Vladimír (referee) ; Dostál,, Petr (referee) ; Pivoňka, Petr (advisor)
Simultaneous evaluation of the whole set of the model parameters of different orders together with an ability to track unmodeled dynamics are desired features in the tasks of parameter estimation. A technique handling with the factors produced by an augmented covariance (ACM) or information (AIM) matrices is considered to be an appropriate tool for designing multiple model estimation. This is where the name augmented identification (AI) by using the least-squares method was taken. The method AI attains numerical stability of the calculation of the conventional least squares method while in the same time, fully extracts information contained in the observation. In order to track time varying parameters can be found that all the information pertinent to recursive identification and thus to data driven forgetting is concentrated in ACM as well as in AIM. In this thesis will be introduced how to selective forgetting to ACM should be applied in an effective way. It means forget only a portion of accumulated information which will be further modified by the newest data included in the regressor. In the estimation problems the knowledge of the inner states of the identified system is often required. Because the augmented identification belongs within the class so called prediction error method (PEM), some rational requirements can be deduced. As a result, state filter should constitute optimization procedure minimizing the predicted error of given state space model representation with respect to the vector of states. The proposed scheme will considerably extend the family of algorithms based on processing of ACM (AIM) about augmented filtering (AF). This all will establish a comprehensive concept of parametric estimation that compared with conventional approaches is characterized by versatility, low demands on a priori process information and by excellent numerical properties (robust against overparametrization, capable solving the multiple model problem).
|
|
|
Closed loop system identification
Piskoř, Dominik ; Buchta, Luděk (referee) ; Pohl, Lukáš (advisor)
This thesis deals with testing of various modifications of the least-squares method, which reduces the bias in system parameter estimation. The first part of this thesis is focused on the theory and principles of every single method. The second part describes the implementation of identification methods in Matlab and Simulink. The third part is focused on the evaluation of used methods via statistical analysis. This part also contains an identification of electric parameters of the real PMSM motor model.
|
|
|
Closed loop system identification
Piskoř, Dominik ; Buchta, Luděk (referee) ; Pohl, Lukáš (advisor)
This thesis deals with testing of various modifications of the least-squares method, which reduces the bias in system parameter estimation. The first part of this thesis is focused on the theory and principles of every single method. The second part describes the implementation of identification methods in Matlab and Simulink. The third part is focused on the evaluation of used methods via statistical analysis. This part also contains an identification of electric parameters of the real PMSM motor model.
|
|
|
Quadratically Optimal Augmented Identification and Filtration
Dokoupil, Jakub ; Bobál, Vladimír (referee) ; Dostál,, Petr (referee) ; Pivoňka, Petr (advisor)
Simultaneous evaluation of the whole set of the model parameters of different orders together with an ability to track unmodeled dynamics are desired features in the tasks of parameter estimation. A technique handling with the factors produced by an augmented covariance (ACM) or information (AIM) matrices is considered to be an appropriate tool for designing multiple model estimation. This is where the name augmented identification (AI) by using the least-squares method was taken. The method AI attains numerical stability of the calculation of the conventional least squares method while in the same time, fully extracts information contained in the observation. In order to track time varying parameters can be found that all the information pertinent to recursive identification and thus to data driven forgetting is concentrated in ACM as well as in AIM. In this thesis will be introduced how to selective forgetting to ACM should be applied in an effective way. It means forget only a portion of accumulated information which will be further modified by the newest data included in the regressor. In the estimation problems the knowledge of the inner states of the identified system is often required. Because the augmented identification belongs within the class so called prediction error method (PEM), some rational requirements can be deduced. As a result, state filter should constitute optimization procedure minimizing the predicted error of given state space model representation with respect to the vector of states. The proposed scheme will considerably extend the family of algorithms based on processing of ACM (AIM) about augmented filtering (AF). This all will establish a comprehensive concept of parametric estimation that compared with conventional approaches is characterized by versatility, low demands on a priori process information and by excellent numerical properties (robust against overparametrization, capable solving the multiple model problem).
|
|
|
Performance, robustness and implementation of controllers
Buchta, Luděk ; Michal, Mrázek (referee) ; Pivoňka, Petr (advisor)
This master‘s thesis deals with the design of modern process control algorithms suitable for SISO industrial control equipments. In this thesis the analysis of the controller design methodology using Mixed-sensitivity function is described. This method consists in shaping frequency characteristics of the sensitivity function and complementary sensitivity fiction using the weighting functions. The proposed H controllers are compared with the classical structure of a discrete PID controller with filtered derivative part. The proposed controllers are compared in terms of robustness, performance, complexity of the design and requirements necessary for their practical application (anti-windup, smooth switchover). Robustness of the controller is evaluated on the basis of modulus stability margin. Estimation of model´s parameters is solved least-square method. The proposed control system consists of an industrial PC and Automation Panel from the company B&R, decentralized system I/O and the real plant. Visualization for control workplace was created in the program Automation studio. This visualization is used to the easy transfer of information from a controlled process.
|
|
|
Adaptive controllers with principles of artificial intelligence and its comparison with classical identifications methods
Dokoupil, Jakub ; Malounek, Petr (referee) ; Pivoňka, Petr (advisor)
This piece of work deals with a philosophy of design adaptive controller, which is based on knowledge of mathematical model controlled plant. This master thesis is focused on closed-loop on-line parametric identification methods. An estimation of model´s parametres is solved by two main concepts: recursive leastsquare algorithms and neural estimators. In case of least-squares algorithm the strategy of preventing the typical problems are solved here. For instance numerical stability, accurecy and restricted forgetting. Back Propagation and Marquardt- Levenberg algorithm were choosen to represent artificial inteligence. There is still a little supermacy on the side of methods based on least-squares algorithm. To compare individual algorithms the grafical interface in MATLAB/Simulink was created.
|
|
| |
guest ::
