Robot Positioning Error Compensation Method Based on Deep Neural Network

• Published in: Journal of physics. Conference series Vol. 1487; no. 1; pp. 12045 - 12054
• Main Authors: Hu, Junshan, Hua, Fangfang, Tian, Wei
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.03.2020
• Subjects: Accuracy; Artificial neural networks; Cartesian coordinates; Error compensation; Error reduction; Hypercubes; Industrial robots; Intelligent manufacturing systems; Latin hypercube sampling; Neural networks; Particle swarm optimization; Physics; Prediction models; Robots
• ISSN: 1742-6588; 1742-6596
• DOI: 10.1088/1742-6596/1487/1/012045

Industrial robots are widely used in intelligent manufacturing industry because of their high efficiency and low cost, but the low absolute positioning accuracy limits their application in the field of high-precision manufacturing. To improve the absolute positioning accuracy of robot and solve the traditional complex error modeling problems, a robot positioning error compensation method based on deep neural network is proposed. The Latin hypercube sampling is carried out in Cartesian space, and the influence rule of target attitude on error is obtained. A positioning error prediction model based on genetic particle swarm optimization and deep neural network (GPSO-DNN) is established to realize the prediction and compensation of the positioning errors. The experimental results show that the positioning error compensation method based on GPSO-DNN presents good compensation accuracy. The positioning error is reduced from 1.529mm before compensation to 0.343mm, and the accuracy is increased by 77.57%. This method can effectively compensate the positioning error of the robot and greatly improve the positioning accuracy of the robot.