A TRAJECTORY TRACKING METHOD OF PARALLEL MANIPULATOR BASED ON KINEMATIC CONTROL ALGORITHM

Shiqi Li,∗ Dong Chen,∗ and Junfeng Wang∗

Keywords

Trajectory tracking, parallel manipulator, dynamic recursive fuzzy neural network, artificial fishswarm algorithm, PID controller

Abstract

A trajectory tracking method based on a dynamic recursive fuzzy neural network (DRFNN) proportional–integral–derivative (PID) controller and an improved artificial fish-swarm algorithm (IAFA) is proposed to control a 6-DOF parallel manipulator. First, a kind of internal feedback structure of universal dynamic neural networks, i.e., the dynamic recursive neural cell, is employed to improve the layers of fuzzy neural network (FNN) and thus to improve the time-varying property and dynamic mapping ability of the network. Then the improved FNN (i.e., the DRFNN) is used to regulate the parameters of the PID controller and thus to track the planned trajectories. Finally IAFA is developed to optimize the parameters of DRFNN and thus to reduce the effect of the initial values of these parameters on tracking accuracy. The proposed method is verified to have good performance both in tracking accuracy and tracking stability.

Important Links:



Go Back