Robust nonlinear fault diagnosis: application to robotic systems

Abstract

Fault detection, diagnosis, and accommodation play an important role in the operation of autonomous robotic systems. Incipient faults in the actuators, sensors or in the dynamics of a robotic manipulator can lead to deterioration in performance and unsafe operating conditions. This paper presents a robust nonlinear fault diagnosis scheme for detecting and approximating faults occurring in a class of nonlinear MIMO systems. The proposed approach utilizes online approximators and adaptive nonlinear filtering techniques to obtain estimates of the fault functions. Performance of the nonlinear fault diagnosis scheme is analyzed by investigating its robustness and fault sensitivity properties in the presence of slowly developing or abrupt faults. A simulation example is presented to illustrate the ability of the fault diagnosis scheme to detect faults in a single link robotic manipulator with a revolute elastic joint.