多目标综合评价中四种确定权重方法的比较与分析

摘要

Currently, the multi-objective comprehensive evaluation method to ascertain the weight of evaluation indexes includ the expert scoring method, principal component analysis ( PCA) , analytic hierarchy process ( AHP) and fuzzy comprehensive evaluation method( FCEM) , etc.The same set of index system, if it is not the same as the method, the corresponding evaluation results will be different.In this paper, we compared research of the four methods of determining the weigh.In our research, 10 drinking water sources as the research object in a city, and choose water consumption per unit of GDP, the unit surface area of TP load and load of COD, nearly comprehensive evaluation.By evaluating the results of the comparative analysis showed that:in the multi-objective comprehensive evaluation, using expert scoring method to determine index weight advantage method is simple, the disadvantage is that the accuracy is lower; using PCA can avoid subjective arbitrariness, most significantly, the differences of performance evaluation index, but the calculation process is more complex and require basic accurate data; the AHP can be objectively to multi-index comprehensive evaluation, the calculation of the weight value distribution region is wider, but is not easy to construct judgment matrix analytic hierarchy process;FCEM is an extension of AHP, on the algorithm is more advantageous to calculation, and the evaluation result is reasonable and reliable.%在多目标综合评价中将指标权重的专家打分法、主成分分析法、层次分析法和模糊综合评价法进行比较研究。以某市10个饮用水源地为研究对象，选取单位GDP用水量、单位水面面积COD负荷、单位水面面积TP负荷等12个评价指标进行综合评价。结果表明：在多目标综合评价中，用专家打分法确定指标权重，其优点是方法操作简便，缺点是精确性比较低；用主成分分析法可避免主观随意性，能显著表现评价指标的差异性，但是计算过程比较复杂且要求基础数据精确；用层次分析法能较客观地对多指标进行综合评价，计算的权重值分布域也更宽，但层次分析法不易构造判断矩阵；模糊综合评价法是层次分析法的一种扩展，在算法上更有利于计算，且评价结果合理可靠。