Abstract:In order to reduce the displacement deformation caused by the temperature effect on the spindle of the milling machine, to improve the machining accuracy of machine, the fuzzy C mean clustering method and multiple linear regression theory are used to model the thermal error of the spindle of the milling machine to realize minimized machining error of the spindle. The iterative process of selecting the optimal value of the fuzzy C mean clustering method was analyzed, the temperature value of different positions on the milling machine was divided into groups, and the optimal temperature value of each group was selected. By using the multiple linear regression theory, the thermal error theoretic prediction model of the milling machine was deduced, and the thermal error prediction model established by the multiple linear regression theory was verified by the experiment. The experimental results show that the temperature effect generated before the compensation, a milling spindle in Y and Z directions of thermal error maximum value respectively are 45.0 μm and 28.0 μm; after compensation, milling machine spindle in Y and Z directions by temperature thermal error maximum value respectively are 3.2 μm and 3.8 μm, and error bounds are less than 4 μm. Using fuzzy C means clustering method and multiple linear regression theory to compensate the thermal error of the milling machine, the error caused by the temperature influence on the main spindle of the milling machine is obviously reduced, thus improving the positioning accuracy of the spindle.