Neural adaptive control for robots with uncertainties in manipulator dynamics and actuator dynamics under constrained task space

Abstract

Driven by the needs of safe human-robot interaction, this paper presents control designs for tracking control of robot system with uncertain manipulator dynamics and joint actuator dynamics subject to constrained task space. The integral Barrier Lyapunov Functional (iBLF) is employed to guarantee the constraint satisfaction, in which a conservative mapping of original constraint into new constraints on tracking errors is removed. Neural Networks (NN) are adopted to approximate the unknown packaged terms in desired control. Appropriate adapting parameters are constructed to estimate the unknown bounds on NN approximation. It is proved that the task space vector of end-effector converge to a small neighborhood around a desired trajectory without violating the predefined constrained region. Simulation results are provided to illustrate the performance of proposed control.