Automatic identification of controlled objects

The paper proposes the new method of identification of a control object by a transient process and investigates it wth modeling. The main idea of the method is the use of transient processes to find the values of the coefficients of the estimated mathematical model of the object. Automatic processes based on the optimization which has its coal to provide minimizing of the cost function. This function is integral of the module or the square of the deviation of the process curve of the model from the actual process curve. In the case of coincidence of the structure of the identified model with the actual structure of the object model, the method is extremely effective. The situation when the proposed structure of the object model differs from the actual one is investigated too. Mathematical modeling showed that in this case the quadratic criterion is less effective than the criterion based on the module of the error. The results of the application of the regulator calculated on the basis of the model obtained as a result of the identification process to the actual object were also investigated. The paper has shown that even with small identification errors, the differences can be large, but using the method of constraining the system frequency band, these models allow calculation of the proper regulator and use automatic identification procedure. Graphs of transient processes are given to illustrate the proposed method.