Module System Monitoring and Fault Detection for Modular and Reconfigurable Robot

Abstract

A distributed module system monitoring and fault detection of the overall joint for modular and reconfigurable robot (MRR) is proposed in this paper. In the design procedure, the joint commanded torque generated by a distributed fault-tolerant controller is filtered and compared with a filtered torque estimate derived from the nonlinear dynamic model of MRR based on joint torque sensing. Considering the motor output capacity, a novel performance index for providing the operation health conditions of an MRR joint is constructed by using the torque prediction error. The proposed scheme is shown to be able to monitor the health of the overall joint for MRR with any movement, which is essential for robots working in human environments. The presented method has been experimentally tested, and the experimental results have demonstrated the effectiveness of the proposed method.