基于参数非参数融合方法的精密机床性能评估  

Performance assessment of precision machine tool based on parameter and non-parameter fusion method

作　　者：田子欣[1] 徐永智[1] TIAN Zi-xin;XU Yong-zhi(Department of Automotive,Sanmenxia Polytechnic,Sanmenxia 472000,China)

机构地区：[1]三门峡职业技术学院汽车学院,河南三门峡472000

出　　处：《机电工程》2022年第11期1614-1619,共6页Journal of Mechanical & Electrical Engineering

基　　金：河南省科技攻关项目(202102210279);三门峡市科技攻关项目(2020010124);三门峡职业技术学院校企合作项目(SZYhxkt-2021-002);三门峡职业技术学院博士专项基金资助项目(SZYGCCRC-2021-007)。

摘　　要：精密机床机械系统误差受传动系统的设计和调试,以及环境变化等多种因素的影响,属于分布未知的乏信息系统,为了很好地解决精密机床机械系统误差数据分布未知的乏信息评估问题,提出了一种用于评估精密机床机械系统性能变异的参数非参数融合方法。首先,在服役现场,至少采集两个不同时间阶段精密机床加工产品尺寸的序列数据,构成了评估对象;其次,采用中位数与Huber M估计融合方法,对不同时间阶段序列数据进行了稳健处理,得到了次序统计量与尺寸性能区间,确定了置信水平;计算出了尺寸性能区间的分组数及尺寸数据列的累积频率,并用斯米尔诺夫假设检验标准统计量,对其进行了对比,评估了不同时间阶段序列数据的分布相似性特征,以此判断精密机床性能是否发生变异;最后,利用专用滚动轴承外圈检测装置,实测两个时间阶段的数据,对该方法的有效性进行了验证。研究结果表明:两个时间阶段产品精度分布是置信水平为95%的偏态分布,二者分布特征不相似,可推断出精密机床机械系统发生了变异;系统通过运用中位数与Huber M融合方法、斯米尔诺夫假设检验方法可以克服数据未知分布及置信水平特征的问题,有效地分析机械系统误差,为精密机床性能评估与维修提供可靠依据。The mechanical system error of precision machine tools is affected by various factors such as transmission system design,debugging,environmental changes,etc.,and belongs to the lack of information system with unknown distribution.Aiming at the evaluation problems of the mechanical system error of precision machine which belonged to poor information of unknown distribution,the parameter and non-parameter fusion method was proposed to assess the performance variation of mechanical system on precision machine.The first of all,in the service time,assessment objectives were constituted by taking at least two order data of products size on precision machine at the different time phase.Then,robust treatment for the assessment objective was conducted by adopting the fusion method of the median and Huber M estimate,order statistics and size characteristics interval were obtained and significance level was determined.Calculated grouping number and cumulative frequency,similarity characteristics of the difference time phases data were evaluated by taking Smirnoff hypothesis test method conducted comparison with the standard statistics.So whether performance variation or not on the precision machine tool were judged.Finally,effectiveness of the methods were tested by precision data of the using two groups data of the special outer rings measuring devices on the bearings at the different time phase.Results show that precision distributions of the 2-time phases are the skew distribution of the 95%significance level,both distribution characteristics is not similarity,so the mechanical system on precision machine have been changed.Unknown distribution and confidence level characteristics of test data can be conquered by systematic using the fusion method of the median and Huber M estimate,Smirnoff hypothesis test,error of mechanical system can be effective analyzed on the above methods.A reliable basis for the performance assessment and maintenance of the precision machine tool complex system is provided.

关 键 词：精密机床 机械系统误差调整 传动系统 分布未知乏信息系统 中位数估计 置信水平 斯米尔诺夫假设 

分 类 号：TH161[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]