Robust Iterative Learning Predictive Fault-tolerant Switching Control for Batch Processes: an Lyapunov-Razumikhin Approach

In the context of batch processes characterized by uncertainty, disturbance, small time delays, and partial actuator faults, a novel robust iterative learning predictive fault-tolerant switching control method is proposed. Building upon the Lyapunov-Razumikhin function, sufficient conditions ensuring system stability are derived and presented in the form of robust positively invariant sets and terminal constraint sets. Under the scenario of small time delays, compared to the conventional Lyapunov-Krasovskii approach, the resulting stability conditions exhibit reduced computational complexity and conservatism. Furthermore, to address the issue of faults, we have devised a switching controller strategy. Under normal conditions, the normal controller is employed, while in the presence of faults, a fault-tolerant controller is activated. In contrast to traditional fault-tolerant control methods, the designed iterative learning predictive fault-tolerant switching controller not only reduces conservatism during regular operations but also ensures robust stability of the system during fault scenarios. The feasibility and effectiveness of the developed method is finally verified with a simulation example.