Study on optimization of temperature measuring points for machine tools based on Grey Correlation and Kohonen network

Abstract

Thermal error modeling is an effective approach to ensure the machining accuracy of NC machine tool by which the thermal error compensation could be implemented. The accuracy and robustness of the thermal error model are affected by the arrangement and identification of the temperature measuring points directly. This paper presents a new approach combining Grey Correlation and Kohonen network to optimize the temperature measuring points of machine tool. All temperatures from measuring points of the machine and thermal deformation of the workpiece are employed as analysis data, and then their gray correlation grade is calculated to obtain their associated sequence and determine the correlation between them, by which the temperature variables are optimized. After that, the thermal key points are selected with correlation coefficient by combining the temperature variables clustered based on the Kohonen network‥ A robust regression thermal error prediction model is presented based on the optimal thermal points, which could reveal the relationship between the measured temperatures and the thermal deformations. The corresponding thermal error prediction experiment was carried out. The predictive accuracy of the proposed thermal error model was compared with that acquired from the temperature key points determined using variable grouping optimization, and the results show that the thermal error model is more accurate by which the prediction accuracy of thermal deformation along X axis and Y axis is greatly increased, which verified the effectiveness and feasibility of the proposed approach, so it is helpful to improve machining accuracy of NC machine tool.