| Abstract [eng] |
The typical optimization methods are analysed in work, which are used for optimization of technological parameters in the control systems. Selected algorithms are classified in two main groups: one-dimensional and multidimensional optimization methods. The algorithms are created in MATLAB/Simulink environment using main functions of the program. Created algorithms enable to find extreme values of mathematical functions and best parameters (for example: integral minimal quadratic error (ISE), integral time-weighted absolute error (ITAE)) of controller as well. Optimization algorithms are designed in order to introduce students with the structure and working principle of the methods. There are many comments in MATLAB/Simulink programs that help to understand programs easily. Selected methods of optimization: Sequential search, Dichotomous method, Golden section search method, Chemotaxis method, Gaus-Zaidel method, Evolutionary programming, Simulated annealing method, Gradient steepest descend method. |
