机电产品性能分析指标体系模型的构建研究

Selection model of performance analysis index system of mechanical and electrical products

针对机电产品性能分析中指标体系的构建问题,对机电产品指标的特征属性和相关性进行了研究,构建指标选取模型,提出了一种基于聚类和因子分析的机电产品指标选取方法。首先根据指标属性进行了特征层分类,构建了指标体系特征层,在特征层下基于聚类分析将信息量相似的指标聚为一类,保证所包含信息既不重复又能覆盖各个方面;然后通过因子分析法进行了指标遴选,确保所选指标对分析结果影响最大;最后将该模型应用在散热翅片性能分析的指标体系构建中。研究结果表明:该指标选取模型将指标数量精简至原60%,同时又包含了95. 4%的信息量,表明该指标选取方法是可行和有效的。

Aiming at the problem of constructing the index system in the function analysis of electromechanical products,the characteristics and correlations of the indicators were studied,an index selection model was constructed,and a selection method of electromechanical products indicators based on clustering and factor analysis was proposed. Firstly,the feature layers were classified according to the attributes of the indicators,and the feature layer of index system was constructed,the indicators with similar information amount were clustered into one category,to ensure that the contained information was not overlapped and can cover all aspects. Then index was selected by factor analysis to ensure that the selected indicators have the greatest impact on the index system. Finally,this model was tested by constructing of index system for performance analysis of cooling fins. The results show that the index selection model reduces the number of indicators to 60%,and includes 95. 4% of information,indicating that the method was feasible and effective.