A simplified LVI-based primal-dual neural network for repetitive motion planning of PA10 robot manipulator starting from different initial states

Abstract

This paper presents a simplified primal-dual neural network based on linear variational inequalities (LVI) for online repetitive motion planning of PA10 robot manipulator. To do this, a drift-free criterion is exploited in the form of a quadratic function. In addition, the repetitive-motion-planning scheme could incorporate the joint limits and joint velocity limits simultaneously. Such a scheme is finally reformulated as a time-varying quadratic program (QP). As a QP real-time solver, the simplified LVI-based primal-dual neural network (LVI-PDNN) is designed based on the QP-LVI conversion and Karush-Kuhn-Tucker (KKT) conditions. It has a simple piecewise-linear dynamics and could globally exponentially converge to the optimal solution of strictly-convex quadratic-programs. The simplified LVI-PDNN model is simulated based on PA10 robot arm, and simulation results show the effective remedy of the joint angle drift problem of PA10 robot.