An Adaptive Fuzzy Logic Approach for Fault Detection in Robot Manipulators with Parametric Uncertainty

A high degree of automation in flexible production units demands powerful tools for supervision and fault detection. This paper outlines recent advances on the theory of model-based fault diagnosis in dynamic systems. Robot manipulator fault detection and diagnosis involves processing of huge information about the robot system. A fuzzy logic based threshold for residual evaluation is suggested here. The proposed fuzzy logic control scheme has been applied to trajectory control of a five-degree of freedom robot manipulator. The proposed method is capable to address unstructured and unknown disturbances. Simulated results are presented in terms of fault detection accuracy and knowledge extraction feasibility.