DESIGN OF AN ADAPTIVE PID CONTROLLER FOR A CABLE-DRIVEN PARALLEL ROBOT BASED ON NEURAL NETWORKS

Abstract

The purpose of this research is to improve the controller performance of a cable-driven parallel robot (CDPR) using a nonlinear PID controller. By combining a linear PID controller with a neural network, which has strong learning, adaptation, and nonlinear problem-solving capabilities, a novel adaptive PID controller is developed. Supervised learning of the neural network is used to minimize a cost function. The proposed controller is suitable for cable-driven parallel robots with nonlinearities and uncertainties. Detailed specifications of the control structure, design process, and experimental results are presented. The experimental results show that the proposed control method achieves higher performance for CDPR in joint space.