Abstract:We investigate the best performance for linear feedback control systems in the case that plant uncertainty is to be considered.First,we define an average integral square criterion of tracking error over a class of stochastic model errors.By utilizing spectral factorization to minimize the performance index,we derive an optimal controller design method and further study best performance in the presence of stochastic perturbation.The results can be used to evaluate optimal performance in practical control system designs.

Best Performance Analysis for Stochastic Perturbation Systems with Spectral Factorization