Adaptive multi-class controller

Abstract

In many process control applications, a single controller design algorithm can not give optimal performance for a wide range of plant structures and operating conditions. In the paper, an adaptive multi-class controller is introduced. Two online identification algorithms run in parallel in the proposed controller and the one with least error cost function value is selected. According to the parameters and structure of the selected estimated model, a controller selection rule-base is developed. Experimental studies are included to demonstrate the performance of the proposed multi-class controller.