A Comprehensive Positioning Accuracy Compensation Method Based on BP Neural Network of Industrial Robots

Abstract

Aiming at the problem that the absolute positioning accuracy of industrial robots cannot meet the requirements of high-precision positioning, a comprehensive positioning accuracy compensation method based on back propagation (BP) neural network was proposed, which considers both the geometric parameters factors and the stiffness performance factors influencing the absolute positioning accuracy of robots. This method uses the actual positioning coordinates and the stiffness performance evaluation index of an industrial robot as the input, and the theoretical positioning coordinates of the robot as output to train a BP neural network. Then the trained BP neural network is used to compensate the absolute positioning accuracy of the robot. This method was tested on a KUKA KR500L340-2 industrial robot, and the experimental results show that the absolute positioning accuracy of the robot is increased from 1.155-2.892mm before compensation to 0.068-0.465mm after compensation. The absolute positioning accuracy of the robot has been significantly improved.