An iterative tuning approach for feedforward control of parallel manipulators by considering joint couplings

Abstract

Feedforward control is an effective way to improve the joint tracking accuracy of robotic systems. This paper presents an iterative approach for feedforward controller parameter tuning of parallel manipulators that considers joint couplings (cross-talk). Based upon a compound control strategy, increments of the feedforward tuning parameters are iteratively updated by minimizing the sum of squares of joint tracking errors at each step. A plant-free identification Jacobian is formulated using the measured data associated with a number of sequential statuses within each iteration cycle. Experiments on the 3-DOF parallel mechanism within a 5-DOF hybrid robot verify parameter convergence and the extrapolation capability of the proposed approach. Compared to otherwise similar feedforward control not considering joint couplings, the root mean square of joint tracking errors was reduced by up to 22% when the mechanism moved at high speed along a path in the neighborhood of the reference configuration.