Abstract: To solve the problem that the existing thermal error modeling methods rely on experience to select temperature measurement points,which leads to poor robustness of the model,a new method of axial thermal error modeling for machine tool spindle based on cutting force was proposed.The influence mechanism of the axial thermal error of NC machine tool spindle on the undeformed chip morphology was studied,the cutting force model considering the axial thermal error of machine tool spindle was established,and the evolution law of cutting force caused by thermal error under the same cutting process parameters was revealed.Then,the relationship between the axial thermal error of machine tool spindle and cutting force was analyzed,and the axial thermal error model of machine tool spindle based on cutting force was established by using support vector regression.Compared with the existing thermal error models,the proposed new thermal error modeling method avoided the complex process of determining temperature sensitive points and a large number of sensors,and did not need to consider the impact of dynamic changes of temperature sensitive points on the robustness of the model.Experiments were carried out in different environments and working conditions,and the prediction accuracy of the model reached more than 90% with strong prediction ability.It provided a new method for on-line identification of machine tool axial thermal error,and provided a technical basis for on-line error compensation and part quality improvement in intelligent manufacturing process.

Key words: thermal error modeling, cutting force, support vector regression, CNC machine tools

摘要： 为解决现有热误差建模方法依靠经验选择温度测量点,导致模型稳健性差等问题,提出一种基于切削力的机床主轴轴向热误差建模新方法。研究数控机床主轴轴向热误差对未变形切屑形貌的影响机制,建立考虑机床主轴轴向热误差的切削力模型,揭示相同切削工艺参数下的热误差致切削力演变规律;分析机床主轴轴向热误差与切削力的关联关系,利用支持向量回归建立基于切削力的机床主轴轴向热误差模型。与传统热误差模型相比,基于切削力的热误差建模方法不需要大量的温度传感器和复杂的温度敏感点确定过程,而且不用考虑温度敏感点动态变化对模型稳健性的影响。通过不同环境和工况下的实验证明,模型预测精度达到90%以上,具有较强的预测能力与泛化能力,为机床轴向热误差在线辨识提供了新方法,同时为机床在线误差补偿以及在智能制造过程中提升零件质量提供了技术基础。

关键词: 热误差建模, 切削力, 支持向量回归, 数控机床