A Systematic Optimization Design of New Constrained Robust Model Predictive Fault-Tolerant Control for Industrial Processes

Abstract

A comprehensive optimization design method of new constrained robust model predictive fault-tolerant control is proposed for industrial processes with partial actuator failures and unknown disturbances. A new state space model which is composed of differential state variables, output tracking errors, and new state variables related to output tracking errors is first constructed according to the multiple single-input single-output (SISO) models. Considering the uncertainties existing in the system, the reproduced model is represented by a discrete switched model with polytopic uncertainties. Subsequently, the model predictive controller with min-max optimization is designed to ensure the performance improvement and solve the drawback that the controller gain is not adjustable. To further enhance the system performance, the designed controller is optimized in two steps based on the switching strategy. In addition, the corresponding switching strategy is designed to realize the mutual switching between different models of the system. Finally, taking continuous stirred tank and injection molding processes as examples, the effectiveness of the proposed method is verified by comparing with the traditional control method.