Topology reconstruction based fault identification for uncertain multi-agent systems with application to multi-axis motion control system.

In the industrial multi-agent systems, the effect of the faults occurring on any subsystem may diffuse to the whole system through network topology and cause the control performance degradation. This paper aims to solve the issue of fault identification for uncertain multi-agent systems with sensor faults. Firstly, a randomized cooperative fault detection system composed by a group of decoupled filters is presented, where a joint fault detection performance assessment scheme is designed. On this basis, a novel topology reconstruction-based fault isolation strategy is proposed. Finally, two case studies on the multi-axis motion control system are conducted, and the results illustrate the effectiveness of the proposed framework. [ABSTRACT FROM AUTHOR]