Control, monitoring and reconfiguration of sampled-data hybrid process systems with actuator faults

This work presents an integrated model-based framework for control, fault detection and control system reconfiguration of hybrid process systems with measurement sampling rate constraints and actuator faults. A family of output feedback controllers are initially synthesized to stabilize each fault-free subsystem with the aid of a dynamic inter-sample model predictor for each mode. The stability properties for each closed-loop subsystem are then analyzed to obtain the maximum allowable sampling period together with an explicit characterization of the fault-free behavior of each mode. Conditions that guarantee asymptotic stability of the overall switched system are also derived and used to examine the interplay between the selection of the sampling period, the model, the controller and observer design parameters, and the dwell time for each mode. To detect actuator faults within a given mode, a time-varying alarm threshold based on the fault-free behavior is obtained and used, and when faults are detected, actuator reconfiguration is performed to maintain closed-loop stability. A key idea of the reconfiguration strategy is to take into account not only the stability properties of the current mode, but also the stabilizing ability and availability of the fall-back actuator configurations for the future modes. The design and implementation of the developed methodology are demonstrated using a hybrid chemical reactor example.