Dynamic Neural Networks for Kinematic Redundancy Resolution of Parallel Stewart Platforms

Redundancy resolution is a critical problem in the control of parallel Stewart platform. The redundancy endows us with extra design degree to improve system performance. In this paper, the kinematic control problem of Stewart platforms is formulated to a constrained quadratic programming. The Karush-Kuhn-Tucker conditions of the problem is obtained by considering the problem in its dual space, and then a dynamic neural network is designed to solve the optimization problem recurrently. Theoretical analysis reveals the global convergence of the employed dynamic neural network to the optimal solution in terms of the defined criteria. Simulation results verify the effectiveness in the tracking control of the Stewart platform for dynamic motions.

I. Introduction

Kinematically redundant manipulators are referred to those which have more degrees-of-freedom (DOFs) than required to determine the position and orientation. The redundancy of parallel manipulators can be utilized to overcome the obstacles, singularities [1], increasing workspace, improving dexterity and to optimize the performance and to achieve a smooth end-effector motion task [2]. As redundant robots have more DOF than required, there usually exist multiple solutions for kinematic control, which motivates us for the consideration of exploiting the extra DOFs to improve the control performance.