Identification of Real Systems using Parameter Perturbed Models

This thesis discusses a method of determining the parameters of a process whose form alone is known. The parameters are identified by constructing an electrical model of the process whose parameters are automatically adjusted until the performance of the model matches that of the system. The model is then an identification of the process. The method of automatically adjusting the model parameters is the parameter perturbation technique in which the parameters are sinusoidally disturbed and the resulting change in model performance is observed and correlated with the original disturbance to obtain information which is used to determine the adjustment of the model parameters. The theory of this technique has been discussed in the literature; this thesis is directed toward the development of effective design techniques and performance criteria for a parameter perturbed model type identifier which will identify a real operating system. An operational identifier was developed which determined the parameters of an X-Y plotter used as an unknown process