适应性FCM聚类分析对导轨温度特征的改进研究

Improvement Research on Temperature Characteristics of Guideway by Adaptive FCM Clustering Analysis

为能够准确地分析导轨的温度特征,采用适应性FCM聚类分析法对导轨温度测点进行合理改进研究,其改进依据为导轨温度及热变形量,通过设置聚类数C的适应性目标分析函数,建立适应性FCM聚类算法可靠性仿真模型,得到温度测点收敛性与鲁棒性较好的多元回归关键测点热误差样本。研究表明,采用对机床导轨预先布置温度测点,通过适应性FCM聚类分析法对温度测点进行合理优化,可将测点数由5个缩减至3个,同时提高了模型预测的准确性与鲁棒性。

In order to accurately analyze the temperature characteristics of the guide rail,an adaptive FCM cluster analysis method is proposed to conduct a reasonable improvement study on the temperature measurement points of the guide rail.Its improvement is based on the temperature and thermal deformation of the guide rail.By setting the adaptive objective analysis function of the clustering number C,the reliability simulation model of the adaptive FCM clustering algorithm is established,and the thermal error samples of the key measurement points of multiple regression with good temperature measurement point convergence and robustness are obtained.The research shows that the temperature measurement points are pre-arranged on the machine tool guide rails,and the temperature measurement points are rationally optimized by the adaptive FCM cluster analysis method,thereby reducing the number of measurement points from 5 to 3.At the same time,the prediction accuracy and robustness of the model are improved.