Research on BP Neural Network Prediction of Position Error Considering the Variation of Industrial Robot Center of Mass

The position error of industrial robots has nonlinear characteristics in the workspace. This paper proposes a back propagation (BP) neural network prediction method for robot position error considering the variation of industrial robot center of mass (CoM). First, the CoM model of the industrial robot is established, which forms a mapping relationship between the joint angle of the robot and the position of the CoM, and the simulation verifies the correctness of the model. Then, the robot position error data set is obtained based on different regional spatial grid sampling methods, and the correlation between the robot end position and CoM about position error is analyzed. Finally, the BP neural network is used to establish position error prediction model that combines the robot end position and the CoM variation. Experimental results show that the constructed model can more effectively predict the robot position error.