Fixed-time Adaptive Neural Control for Robot Manipulators with Input Saturation and Disturbance

A fixed-time adaptive neural network control scheme is designed for an unknown model manipulator system with input saturation and external environment disturbance, so that the system convergence time can be parameterized and not affected by the initial state of the system. The compensation control item is introduced to compensate for external disturbance. The scheme can ensure that the input torque always does not exceed the actuator saturation value and the transient and steady state performance will not significantly degrade. Furthermore, the Incremental Broad Neural Network (IBNN) is used for approximating unknown models with flexible adjustability and high computational efficiency, so it can be applied to scenarios with different control precision requirements. Simulation results verify the effectiveness of the scheme in the above aspects.