摘要
将灰关联聚类分析法用于CJ20系列接触器动态性能评价,以9台CJ20-25型交流接触器为评价对象,实现动态性能的等级归属和优劣排序2种评价。提出了针对"样本-相角-指标采样值序列"三维数据的算法步骤和公式,分别采用区间数和实数2种类型进行决策评价,并分析了这2种数据处理结果存在差异的原因。分析表明:1)灰关联聚类分析法用于评价动态性能可同时实现等级归属评价和性能优劣评价;2)考虑全相角范围进行性能评价时,区间数算法比结合统计学实数的算法更易受极值影响;3)若研判特定相角时的性能,鉴于目前尚无公认统一的区间数排序方法,故结合统计学的实数算法的评价结果更合理。
Grey-correlation clustering method was firstly applied to evaluate the dynamic characteristics of CJ20 series contactor.Nine CJ20-25 AC contactors were chosen to rank and to sort their dynamic performance.This paper proposed the steps and formula of the algorithm in detail which aimed to ＂sample-phase-sampled value sequence＂.Two data forms included interval and real numbers were used in the method.And the causes of difference between the two forms were analyzed.The analysis showed that：1） the grey-correlation clustering method could be used to rank and sort at the same time;2） if taking the whole angles into account the interval number algorithm would be more susceptible by extreme data than real number one;3） if taking some specified angle into account,in view of the fact that until now there was still no certain and correct method how to rank the interval numbers,so the real number algorithm combined with statistics was more reasonable.
出处
《南昌大学学报（工科版）》
CAS
2017年第4期393-397,共5页
Journal of Nanchang University(Engineering & Technology)
基金
国家自然科学基金资助项目(51277031)
福建省科技厅重点资助项目(2014H0002)
河北省教育厅科技项目(2017040)
关键词
CJ20系列接触器
动态性能评价
区间数
灰色关联聚类法
CJ20 series contactor
dynamic characteristic evaluation
interval number
grey-correlation ＆clustering method
