Actuator fault-tolerance evaluation of linear constrained model predictive control using zonotope-based set computations

Abstract

This paper presents a computational procedure to evaluate the fault tolerance of a linear-constrained model predictive control (LCMPC) scheme for a given actuator fault configuration (AFC). Faults in actuators cause changes in the constraints related to control signals (inputs), which in turn modify the set of MPC feasible solutions. This fact may result in an empty set of admissible solutions for a given control objective. Therefore, the admissibility of the control law facing actuator faults can be determined by knowing the set of feasible solutions. One of the aims of this paper is to provide methods to compute this set and to evaluate the admissibility of the control law for a given AFC, once the control objective and the admissibility criteria have been established. In particular, the admissible solution set for the predictive control problem, including the effect of faults (either through reconfiguration or accommodation), is determined using an algorithm that is implemented using set computations based on zonotopes. Finally, the proposed method is tested on a real application consisting of a part of the Barcelona sewer network.