RNN Based Adaptive Compliance Control for Robots with Model Uncertainties

You are here: Homepage > Search

RNN Based Adaptive Compliance Control for Robots with Model Uncertainties

Free access

Zhou, Xuefeng / Xu, Zhihao / Li, Shuai / Wu, Hongmin / Cheng, Taobo / Lv, Xiaojing

Position-force control is challenging for redundant manipulators, especially for the ones considering both joint physical limitations and model uncertainties. In this chapter, we considered adaptive motion-force control of redundant manipulators with uncertainties of the interaction model and physical parameters. The whole control problem is formulated as a QP equation with a set of equality and inequality constraints, where based on admittance control strategy, the desired motion-force task is combined with the kinematic property of redundant manipulators, corresponding to an equality constraint in the formed QP equation. Moreover, the uncertainties of both system model and physical parameters are also considered, together with the complicated joint physical structure constraints, formulating as a set of inequality constraints. Then an adaptive recurrent neural network is designed to solve the QP problem online. This control scheme generalizes recurrent neural network based kinematic control of manipulators to that of position-force control, which opens a new avenue to shift position-force control of manipulators from pure control perspective to cross design with both convergence and optimality consideration. Numerical results on a 7-DOF manipulator LBR iiwa and comparisons with existing methods show the validity of the proposed control method.