Performance assessment of control loop with multiple time-variant disturbances based on Multi-model Mixing Time-variant Minimum Variance control

摘要

The disturbance models in industry system always change with several disturbances existing simultaneously. The performance assessment based on Single Model Time-variant Minimum Variance (SMTMV) is not available to system with multiple time-variant disturbances. Since the transient deviation exists, the performance assessment based on Multi-model Switching Time-variant Minimum Variance (MSTMV) is not always available, too. In order to solve this problem, the method of performance assessment based on Multi-model Mixing Time-variant Minimum Variance (MMTMV) is proposed. First, we design SMTMV controllers for every time point according to the time-variant disturbance models. Then we obtain MMTMV controller by the multi-model mixing theory using the SMTMV controllers and the variance of the MMTMV controller's output is the benchmark to the performance assessment of control loop with multiple time-variant disturbances. Simulation results indicate that the method of MMTMV control is available.