Guaranteed performance adaptive fixed-time neural control for robot manipulators with disturbance observer

This paper studies the fixed-time tracking control scheme for robot manipulators within the constraints of prescribed performance bounds (PPB). Firstly, a disturbance observer with exponential convergence performance is developed to estimate the lumped disturbances and actuator failures, a radial basis function neural network (RBFNN) is introduced to reduce the estimation error which generated by the observer. Based on prescribed performance function, an adaptive fixed-time fault-tolerant control (AFFTC) is developed using the back-stepping control framework. This approach does not necessitate complete information about the robot manipulator’s dynamic model, thereby making its implementation in practical, real-life situations more accessible. The stability of the proposed system is established via the Lyapunov criterion and the feasibility of the observer is validated.