A Fault Tolerant Control for Robotic Manipulators Using Adaptive Non-singular Fast Terminal Sliding Mode Control Based on Neural Third Order Sliding Mode Observer

Abstract

This paper proposes a fault tolerant control technique for uncertain faulty robotic manipulators when only position measurement is available. First, a neural third-order sliding mode observer is utilized to approximate the system velocities, the lumped uncertainties and faults, in which the radial basis function neural network is employed to approximate the observer gains. Then, the obtained information is applied to design a non-singular fast terminal sliding mode control to deal with the effect of the lumped uncertainties and faults. In addition, an adaptive law is used to approximate the sliding gain in switching control law. The controller-observer method can provide superior features such as high tracking precision, less chattering phenomenon, finite-time convergence, and robustness against the lumped uncertainties and faults without the requirement of its prior knowledge. The stability and finite-time convergence of the proposed technique are proved in theory by using the Lyapunov function. To verify the usefulness of the proposed strategy, computer simulations for a 2-link serial robotic manipulator are performed.