Research On Thermal Error Modeling And Compensation Method Of CNC Machine Tool Spindle

With the improvement of CNC machine tool manufacturing level,thermal error has gradually become the main factor affecting machine tool’s processing accuracy,a large number of studies show that thermal error in the overall error ratio has reached40%～70%.As the main moving part and heat generating part of machine tool,the research on thermal error of spindle system is always the key and difficult point of NC machining research.In order to improve the machining accuracy of machine tools,this paper takes DSX400 CNC machine tool spindle system as the research object,and establishes a thermal error prediction model based on particle swarm optimization(PSO-SVM)under variable working conditions.Through comparative experiments,the results show that PSO-SVM model has higher prediction accuracy.Later,according to the model of the spindle of the machine tool compensation experiment,the compensation system can reduce 77.79% thermal error.Machine tool machining accuracy has been greatly improved after compensation.The main research contents are as follows:In terms of the selection of optimal temperature measuring points,a method based on the combination of finite element analysis,fuzzy clustering and grey correlation degree is proposed.According to the thermal deformation of finite element analysis,the location of temperature measuring points is preliminarily determined.Then the method combining fuzzy clustering and grey correlation degree is used to get five optimal temperature measuring points of spindle system.The temperature of the five optimal temperature measuring points can better reflect the overall temperature of the spindle system,and the model accuracy can be improved while reducing the number of sensors.In terms of the selection of thermal error model,under the circumstance of variable condition,a model was proposed based on particle swarm optimization support vector machine(SVM)model,combining the searching capability of the PSO algorithm and support vector machine(SVM),the advantages of the Z to PSO-SVM model.After comparing the prediction made by support vector machine(SVM),multiple linear regression(MLR),the results indicate that the prediction accuracy of this model is higher under variable working conditions.In terms of thermal error compensation,a thermal error compensation system was designed.Based on the PSO-SVM model and combined with the origin offset compensation method and semi-closed loop feed-forward control system,the system compared the mean and root mean square values of thermal errors in the X and Z directions before and after compensation,and the error range of the workpiece contour before and after compensation.The results show that the machining precision is higher after compensation.