Automatic Regulating Time Series for Multivariable Processes with Specifications on Rise Times

Abstract: In this paper, an auto-tuning method is proposed for multivariable processes. The considered system is first estimated by use of an on-line identification method, and then low-interacting precompensating matrix is applied , and finally nonlinear automatic regulating time series is applied to get the wanted response times. The performance of the compensated systems can be evaluated by wanted rising time and low-interacting characteristics. Three numerical examples give the proposed method provides better performance and less interaction effects than those of other four mentioned on-line computing and PI controlling methods.