Event-Triggered Fault-Tolerant Control of Sampled-Data Nonlinear Systems

Abstract

Abstract: This work focuses on the stabilization of a class of uncertain nonlinear systems with sensor-controller communication constraints, sampled measurements and control actuator faults. Initially, an event-triggered model-based control strategy that enforces closed-loop stability under continuously-sampled measurements is developed. The strategy involves using the model state to generate the necessary control action for periods of time when sensor-controller communication is suspended, and updating the model state using the state measurement when communication is restored. Communication is triggered whenever a state-dependent threshold on the model estimation error is breached. The communication threshold is characterized in terms of the model and controller design parameters, as well as the fault severity. A modification of the communication threshold is then presented to address the implementation of the control strategy under discretely-sampled measurements. An on-line forecasting mechanism that provides a bound on the growth of the model estimation error over each sampling interval is used to obtain a tighter communication trigger which guards against potential instability due to the lack of state information between sampling times. Implications of the results for the development of actuator fault accommodation and reconfiguration strategies that guarantee closed-loop stability and keep network resource utilization to a minimum in the presence of faults are discussed. Finally, the results are illustrated using a chemical process example.