Abstract: To eliminate the influence of spindle thermal error on machining quality, smart temperature sensors and displacement sensors were mounted on a precise horizontal machining centre. A temperature-thermal error model of spindle was set up on the assumption that the spindle's temperature field was quasi-static. To optimize data performance and improve dynamic quality of thermal error compensation system, grey theory was used to construct thermal error prediction model. By setting up the metabolic prediction model of temperature, the effect caused by the time difference between temperature detection and error compensation was avoided, and the foresighted prediction of spindle thermal error was realized. The result showed that the on-line foresighted model had a small amount of calculation with high accuracy. The foresighted model could be used in real-time prediction in steady machining process.

Key words: thermal error, spindle, modelling, Foresighted prediction, machine tools

摘要： 为消除主轴热误差对加工质量的影响，以某精密卧式加工中心为研究对象，利用智能温度传感器和位移传感器检测机床温度值和主轴热变形，基于主轴温度场准静态假设建立了主轴温度—热误差模型。为优化数据性能，改善补偿系统动态品质，选取灰色理论建模，通过建立温度新陈代谢预测模型，消除了温度检测和补偿实施的时差影响，最终实现主轴热误差的前瞻预测。研究结果表明，该模型计算量小，预测精度高，可用于稳定加工中的热误差实时预测。

关键词: 热误差, 主轴, 建模, 前瞻预测, 机床