In order to improve the prediction accuracy of the thermal error model and reduce the number of temperature sensors arranged inside the machine tool,a spindle axial thermal error identification model based on the data of a single temperature sensor was proposed.The input of the model was generated by processing the data collected by a single temperature sensor,and there were few internal parameters.The global optimization ability of the intelligent optimization algorithm was used to identify the model parameters,reduce manual intervention,and make the model more generalizable.Taking a certain type of three-axis machine tool as the experimental object,the model identification effect was verified by cutting workpiece with the machine tool.Through comparative analysis and experimental verification with the neural network spindle thermal error prediction model,the results show that the proposed thermal error model has a small residual error in predicting the axial thermal error of the main shaft,high prediction accuracy,and has the advantages of less internal parameters and strong generalization ability,can support the integrated application of CNC machine tools.