1998 (Volume 7)
On some characterization of invariant and decoupling zeros of singular systems
(Military University of Technology, Poland )
In this paper a geometric characterization of invariant and decoupling zeros of a generalized state-space model described by the matrix 5-tuple (E,A,B,C,D), where E is singular but the pencil sE-A is regular, is presented. These zeros are characterized as invariant zeros of an appropriate linear system S(A,B,C,D). Several numerical examples are included to illustrate the proposed results as well as to discuss different definitions of zeros.
Location problems from the multiple criteria perspective: efficient solutions
(Warsaw University, Poland )
Location problems can be considered as multiple criteria models where for each client (spatial unit) there is defined an individual objective function, which measures quality of a location pattern with respect to the client satisfaction (e.g. it expresses the distance or travel time between the client and the assigned facility). The individual objective functions are usually conflicting when optimized. Therefore, the decision maker or planner needs to select some compromise solution for implementation. In this paper we analyze various approaches to discrete multiple facility location problems (various solution concepts) from the perspective of the multiple criteria models. We focus our analysis on two aspects of the solution concepts: if a generated solution is an efficient (Pareto-optimal) solution to the multiple criteria problem, and if the solution concept provides some control parameters allowing the decision maker to select every efficient solution of the multiple criteria problem. That means, we analyze if a solution concept complies with the optimality principle for the multiple criteria model as well as if it allows to take into account various preferences of the decision maker.
keywords: Discrete Location Problem, Multiple Criteria, Efficiency.
Diagnostics of processes in decentralized structures
|Jan Maciej Kościelny|
(Warsaw University of Technology, Poland )
The paper presents rules for diagnosis formulation in hierarchical systems. The rules can be applied in industrial process diagnostics realized in decentralized structures.
Formalized description of an object of diagnosis as well as general rules for diagnosis formulation have been presented in the paper. The method of hierarchical description of the object of diagnosis has been given. Decentralized structures have feedbacks between the subobjects that cause symptoms of faults in one subobject to appear in other ones. Rules for fault isolation in such structures have been presented. Considerations are illustrated with help of examples.
keywords: diagnostics, fault isolation, industrial processes,large-scale systems, decomposition, hierarchical structures.
Optimal design of control law under uncertainty by dynamic programming
|Mariusz Siomak and Krzysztof Malinowski|
(Warsaw University of Technology, Poland )
In this paper various aspects of the realization of the dynamic programming algorithm are presented. Particular attention is focused on computational efficiency, which is obtained by presented discretization, approximation and parallelization techniques. Numerical solutions of some examples are briefly described and discussed. The last part presents a particular approach to handle stochastic constraints with illustrative examples.
All numerical results were obtained on the CRAY 6416 super-server installed in Computing Center of the Warsaw University of Technology.
The central control accuracy theorem for internal model control and its applications
|Krzysztof Latawiec, Czesław Marciak and Janusz Wrzuszczak|
(Technical University of Opole, Poland )
The well-known feature of the Internal Model Control (IMC) structure, providing steady-state error-free servo and regulatory controls for step-wise forcing signals, is generalized by what is termed the "central IMC theorem". The generalization consists not only in the original extension of the SISO and square MIMO results to nonsquare MIMO systems, but mainly in redeveloping certain IMC accuracy-related properties to any appropriately scaled, stabilizing controller. Instructive applications of the central IMC theorem in Model Algorithmic Control (MAC), Extended Horizon Model Algorithmic Control (EHMAC) and Generalized Predictive Control (GPC) are demonstrated.
keywords: internal model control, multivariable control, control accuracy, predictive control, robust control, nonsquare multivariable systems.
An outline of the linear control system synthesis by a proper, stable rational functions approach
(Warsaw University of Technology, Poland )
In the process of designing controllers for linear multivariable plants specially effective are algebraic methods which require from the transfer matrices of both, the plant and the controller to be presented in coprime fractional form with factorization carried on with respect to the ring of exponentially-stable, proper real-rational functions.
The main objective of the paper is to show that this form of representation with simultaneous parametrization of all linear controllers that provide internal stability of the closed-loop system can be achieved in the simplest and most natural way by analysing the system shown in Fig. 3 - the so-called basic structure.
Problems of choosing the parameter to meet some important design specifications, viz. a robust asymptotic tracking of the reference signal with disturbance and noise rejection are also considered and illustrated by two representative examples covering the area of continuous- and discrete-time systems.
Automatic generation control of multi area interconnected power system considering nonlinearity due to governor dead band
|Abani Mohan Panda|
(Indira Gandhi Institute of Technology, India )
The paper focuses on the problem of automatic generation control of multi area power system considering nonlinearity due to governor dead band. The objective of this work is to develop a mathematical model and method of solution to study the dynamic response and to present an optimization method to choose the optimal gain setting of the load frequency controller. A method, called decomposition technique is proposed to study the dynamic response of multi area interconnected power system with AGC. Using this technique a system with large number of interconnected areas with widely different area characteristics can be modeled and studied. Another algorithm is proposed to implement conjugate gradient method for optimization of controller gain parameter by minimizing certain performance index. A case study with four area interconnected system including governor dead band shows that the method is effective.
keywords: Mathematical modeling, Optimization, Automatic generation control, Power systems and plant.
The use of the MTS method for programming of Simatic S7-200 controllers
|Tadeusz Mikulczyński, Zdzisław Samsonowicz and Rafał Więcławek|
(Technical University of Wrocław, Poland )
An outline of the method of net transformation (MTS) designed by the authors for modelling discrete processes and programming PLCs is presented. Characteristics of Simatic S7-200 controller, which is a representative of a new family of Simatic S7 controllers from Siemens, are provided. The use of the MTS for the programming of PLCs is described and illustrated with the modelling of a real (model) process of dosing loose materials and programming Simatic S7-200 controller.
A neuro-fuzzy approach to system modelling Part II. Applications
|Marian B. Gorzałczany|
(Kielce University of Technology, Poland)
The description of behaviour of complex and ill-defined systems and processes is usually based on a combination of two types of knowledge and data: a qualitative, fuzzy knowledge which contains elements of uncertainty and vagueness and often is expressed in the form of linguistic rules usually provided by a domain expert, and a quantitative, nonfuzzy information which appears in the form of measurements and other numerical data. Part I (ACS No. 1/2 1998) of this paper presents a methodology for modelling of complex systems which can effectively represent, process and generalize both mentioned-above types of system's knowledge. The proposed methodology combines artificial neural networks with some elements of the theory of fuzzy sets and fuzzy logic, yielding a structure that can be called a fuzzy neural network or a neuro-fuzzy system. Two examples of the application of our approach in the area of system modelling are presented in Part II of the paper.
keywords: Fuzzy sets, artificial neural networks, neuro-fuzzy systems, fuzzy neural networks, fuzzy models, identification methods, fuzzy controllers.