An infinite horizon model predictive control based active fault tolerant adaptive control scheme

Abstract

This paper proposes an adaptive infinite horizon model predictive control (IHMPC) integrated with active fault tolerance properties due to the performance degradation of the model predictive control (MPC) caused by operational faults. The proposed scheme includes a fault supervision layer composed of fault diagnosis and accommodation methods to provide the ability to update the nominal model of the controller. The simulation results in a nonlinear industrial reactor subject to process faults illustrate that the proposed approach achieves better dynamic performance and has a reduced computational cost compared to a robust IHMPC.