Control system identification through model modulation methods

Abstract

Adaptive control systems techniques have been utilized to investigate identification of the dynamic equations of operating control systems. The identification has been achieved by using model modulation techniques to drive dynamic models into correspondence with operating control systems. The system identification then proceeded from examination of the model and the adaptive loop. The model modulation techniques applied to adaptive control systems are briefly discussed. These techniques applied to simulation studies on an analog computer and a digital computer, and applied directly to a position servo-mechanism, are then discussed. The discussion presents data showing the accuracy of identification of the dominant characteristics of the dynamic equations when the order of the equation is either known or approximated. It was found that these dominant roots could be found within an average accuracy of 5% for both complex phase angle and undamped natural frequency. The authors wish to express their appreciation to Professor Richard C. Dorf of the U.S. Naval Postgraduate School, for the suggestion of the model modulation method and for his assistance during this investigation.