LS-SVM Modeling Based Inverse Controller With Uncertainty Compensation

Abstract

Least squares support vector machine (LS-SVM) modeling based inverse controller (IC) is presented for excitation control of synchronous generator. This IC strategy design includes two main parts: inverse control law and uncertainty compensation. The inverse control law, designed for the control of plant dynamics, is derived directly based on Taylor expansion and it is implemented using LS-SVM modeling. In addition, a robustness filter in the feedback structure, designed for plant disturbance, is employed as uncertainty compensation. The robust stability of the proposed controller is analyzed based on Lyapunov function. Simulations demonstrate the effectiveness of the IC strategy for excitation control.