Active Fault Tolerant Controller Design using Model Predictive Control

Abstract

This paper presents an approach for active fault-tolerant control (FTC) design of constrained linear multi input-multi output (MIMO) systems. The proposed approach is based on model predictive control (MPC) and fault estimation scheme. This method uses two optimal observers to estimate the plant states and loss of effectiveness factor of actuators and sensors. A supervisor unit also uses the fault modelling and correction of plant model per sampling time to accommodate simultaneous partial actuator and sensor faults. Also, by using embedded integrator and feedback compensation in MPC formulation, actuator and sensor faults are compensated automatically. The most important advantage of the proposed approach is low on-line computational load, ability to deal with the constraints and all types of faults in control system simultaneously and its simplicity for real applications. Simulation results on active magnetic bearing system show the effectiveness of the proposed approach.