摘要
轮槽铣床主轴箱热变形数据样本量繁杂,利用传统的FCM模糊聚类算法对主轴箱温度测点进行分析时,需根据情况自行设置分类数,由于经验不足会使分析结果出现偏差,导致分析失效。基于以上情况,提出采用改进的FCM模糊聚类算法对主轴箱测点进行优化分析,其原理为依据主轴箱温度及热变形量,增设聚类数c的自适应目标函数,并建立了改进的FCM模糊聚类算法可靠性分析模型,基于该模型分析得到了多元回归关键测点热误差分析数据。结果显示:采用FCM聚类算法对轮槽铣床主轴箱预先布置的温度测点进行分组优化,使主轴箱的关键测温点由21个缩减至6个,且分析结果准确度较高。该方法为机床温度测点优化分析提出了新的思维路径,具有较好的应用前景。
The sample volume of the thermal deformation data of the spindle box of the wheel groove milling machine is complicated.When the traditional FCM fuzzy clustering algorithm is used to analyze the temperature measurement points of the headstock,the number of categories needs to be set according to the situation.Due to lack of experience,the analysis results will be biased,resulting in analysis failure.Based on the above situation,an improved FCM fuzzy clustering algorithm is proposed to optimize and analyze the measuring points of the headstock.The principle is to add an adaptive objective function of cluster number C according to the temperature and thermal deformation of the headstock,and establish an improved reliability analysis model of FCM fuzzy clustering algorithm.Based on the analysis of the model,the thermal error analysis data of the key measurement points of multiple regression are obtained.The results show that the FCM clustering algorithm is used to group and optimize the temperature measurement points pre-arranged in the spindle box of the wheel groove milling machine,which reduces the key temperature measurement points of the spindle box from 21 to 6,and the analysis results have high accuracy.This method presents a new thinking path for the optimization analysis of machine tool temperature measurement points,and has a good application prospect.
作者
李志伟
LI Zhi-wei(Sichuan College of Architectural Technology,Deyang 618000,China)
出处
《机械工程与自动化》
2020年第3期140-142,共3页
Mechanical Engineering & Automation
关键词
FCM模糊聚类算法
主轴箱
温度测点
优化分析
FCM fuzzy clustering algorithm
spindle box
temperature measurement point
optimization analysis
