Imprementation of Sochastic Searching for Complex Processes Identification

This paper presents implementation of stochastic searching method for complex technological process identification. As the first step of control systems design is identification of a process mathematcal model. The author uses certain stochastic algorithms created & explained in some of his earlier published papers. The basis of the said algorithms is a non-linear stochastic searching. From general math process description this paper considers a process steady state property as an object of identification. Some of presented examples are useful to observe possible problems most of control system designers face through attempts to solve control problems or to modify process technology. As fas as the implemented algorithms are concerned certain advantages & efficiency have been obtained vs regular gradient methods. Since the said algorithms are with some modification useful for training of artificial neural networks, the future development could be toward an inteligent control systems design & implementation. Numerical procedures have been processed by using MATLAB.