Optimal Tracking Cooperative Control for Cyber-Physical Systems: Dynamic Fault-Tolerant Control and Resilient Management

Abstract

This article proposes a novel dynamic fault-tolerant control model to address the optimal tracking cooperative control problem for cyber-physical systems, by considering that all systems can be endowed as a multiagent system and the admissible levels of the actuator fault can be resiliently management. Different from previous works, the feedback gain for the cooperative controller design is no longer fixed, and actuator outage behaviors can be solved by a resilient control way. By introducing a sampling manner, a robust optimal framework is first developed to determine the appropriate feedback gain under a cost constraint for the dynamic fault model. The dynamic fault-tolerant control protocol is, then, designed to achieve the cooperative behaviors. Moreover, a fault management mechanism is proposed, in which the fault parameter is reset as an initial value when the fault growth is greater than the admissible level. By this design, the tracking cooperative behaviors can be achieved in a resilient management process. Two examples are presented to illustrate the effectiveness of the proposed theories.