Robust self-tuning regulator for a synchronous generator

The paper is concerned with an application of a self-tuning control strategy to a synchronous generator system. In the stochastic control strategy, an auxiliary predictor is introduced and the variance between the actual output of the system and the desired track of the output is directly predicted, to restrain excessive control signals and to improve the control performance of the system. The paper also proposes a new supervision scheme for the robust self-tuning regulator, which includes modification of the covariance matrix in the estimation algorithm, moving boundaries imposed on parameter values, and turning on or off the control algorithm to protect estimated parameters from bursting, and to enhance parameter tracking during the dynamic andtransient conditions encountered over the full range of operating conditions. Results obtainedusing a detailed nonlinear simulation of a turbogenerator illustrate the effectiveness of the control strategy and the supervision scheme.