Output feedback robust predictive fault-tolerant hybrid switching control for a nonlinear system

Abstract

Due to the widespread issues of uncertainty, nonlinearity, partial actuator faults, and unmeasurable states in modern production processes, the present study developed an output feedback robust predictive fault-tolerant hybrid switching control method. The nonlinear system is described by a multiphase switching model, which largely restores the system’s nonlinear dynamics. Considering partial actuator faults, the multiphase switching model is divided into normal and fault cases. Based on this model, and to address the issue of an unmeasurable state, a robust predictive fault-tolerant hybrid switching controller is developed involving normal and fault-tolerant controllers, providing effective control in both cases. Subsequently, sufficient conditions based on linear matric inequality forms are provided, which are solved to obtain control law gains, thereby ensuring the system’s stability under both normal and fault cases. In addition, robust stability analysis and exponential stability analysis are performed, which provide the basis for the given sufficient conditions and deliver the dwell time for each phase of the system, respectively. Ultimately, the simulation using a continuous stirring reaction reactor validates the excellence of the suggested approach over traditional fault tolerance and model predictive controls, showcasing enhanced fault tolerance, reduced output and input fluctuations, and improved tracking in normal and faulty conditions.