This paper first analyzes performance indices based on minimum variance control which was proposed by Harris in 1989.With time series analysis and minimum variance control principle as analytic means, we discuss the familiar control architectures in thermal automation systems and educe a control performance assessment algorithm which can be conveniently used in actual projects. We use Marquardt method to estimate the parameters of ARMA model and apply Ftest to determine the orders of ARMA model. Subsequently, we adopt feedback controllerinvariant theory(minimum variance control theory) to acquire the output minimum variance. Then, we can calculate the control performance indices of the control systems. We apply the above assessment technique to feedback control systems, feedforward/feedback control systems, cascade control systems, and gain good effect.
