An intelligent scheme to monitor the status of a power system

A real-time intelligent scheme is proposed to (a) monitor stability, (b) monitor performance in terms of an integral squared error, or unit step response measures such as the degree of damping, rise and settling times, (c) predict overload and (d) predict faults in sensors, actuators or controllers. It consists of (i) a knowledge base and (ii) a linear predictive coding algorithm (LPCA) to estimate in real time the transfer function model of the error signals in the control loops. The error signals, which include the frequency error, voltage error and the area control error, serve as the ‘eyeʹ and ‘earʹ of the intelligent system. The coefficients of the numerator and the denominator polynomials of the transfer function form the linear predictive codes. The proposed scheme is evaluated by (i) a simulated model in real time of a two-area automatic generation control system and (ii) an implementation on a real-time position control system.