该文分析了数控卧式车床热误差对机床加工精度的影响。利用红外热像仪、位移传感器和温度传感器记录热误差数据，建立热误差模型和热误差补偿系统。利用红外热图像和相关分析，对关键测温点的位置进行了优化。然后，建立了车床主轴径向（X方向）热误差的线性回归模型。实验结果表明：线性回归模型是鲁棒的，适用于机床热误差建模。利用线性回归模型开发了基于Siemens828D型数控系统及S7-300PLC（可编程逻辑控制器）的热误差补偿系统。检测结果表明：轴径向热误差由原来的10 μm减少到5 μm以内，精度提高50%以上。

The influence of thermal error on machine tool processing accuracy was analyzed for a computer numerical control (CNC) horizontal lathe. Thermal error data was recorded using an infrared thermal imager, a displacement sensor and a temperature sensor to establish a thermal error model and a thermal error compensation system. Infrared thermal images and correlation analyses were used to optimize the locations of key temperature measurement points. Then, a linear regression thermal error model was developed for the main lathe axis in the radial direction (X direction). Tests show that the linear regression model is robust and suitable for thermal error modelling of the lathe. The linear regression model was then used to develop a thermal error compensation system based on a Siemens 828D CNC system and an S7-300PLC (programmable logic controller). Tests show that the radial thermal error of the shaft is reduced from the original 10 μm to less than 5 μm with the accuracy improved by more than 50%.