Research on Thermal Error Modeling Method of Machine Tool Spindle Based on Optimized BP Neural Network
In view of the limitations of the traditional single temperature measuring point in monitoring the temperature change of the spindle of CNC machine tools,and the shortcomings of the thermal error model based on backpropagation neural network(BP)in accuracy,convergence and robustness,an improved adaptive particle swarm optimization backpropagation neural network(IAPSO-BP)model based on multiple temperature sensors was proposed,aiming to improve the identification accuracy of thermal error of spindle.Multiple temperature sensors were introduced to comprehensively monitor the temperature information of the spindle.The application of adaptive particle swarm optimization could reduce the need of manual parameter adjustment and improve the generalization ability of the model.Taking a specific type of machine tool as an example,the thermal error model of the spindle was established through the real cut-ting experiment,and its validity and robustness were verified.The experimental results show that compared with the traditional BP neural network prediction model,the mean square error of the proposed IAPSO-BP model is reduced by 67.45％,the maximum absolute resid-ual is reduced by 69.62％,and the goodness of fit is increased by 4.29％,which proves the superiority of the model.
machine tool spindleparticle swarm optimization algorithmBP neural networkthermal error model
万方数据
维普
