基于SAGAFCM与主成分-熵的列车开行方案评价指标体系
Evaluation index system of train plan based on SAGAFCM and principal component-entropy

旅客列车开行方案评价指标体系是开行方案评价的基础,评价指标体系的建立和指标的选取直接影响到评价结果的准确性。现有的列车开行方案评价指标体系中,某些指标之间具有较强的相关性,为了避免该类指标对方案的某一方面做重复评价,需要对其进行分类和精简。根据评价指标之间的相互关系,采用基于遗传模拟退火算法的FCM算法对开行方案评价指标体系进行聚类简化,利用主成分分析法和信息熵法对其简化前后的信息量进行测量,得出保留原评价指标体系信息量相对较多的最佳简化评价指标体系。对文献中建立的评价指标体系进行重新简化,测试结果表明,重新构建的简化评价指标体系相较原简化评价指标体系能保留更多的信息量,证明了该方法的有效性。
The evaluation index system of passenger train plan is the basis of the evaluation of train plan. The establishment of evaluation index system and the selection of indexes directly affect the accuracy of the evaluation results. In case of repeated examination on index of one aspect of the program, some indicators that have relevance to another in the existing train plan evaluation index system need to be classified and streamlined. The FCM algorithm based on genetic-simulated annealing algorithm was adopted to cluster and simplify the evaluation index system of the train plan according to the correlation between the evaluation indices. Principal component analysis and information entropy method were used to measure the amount of information before and after the simplified evaluation index system. The relatively good simplified evaluation index system with the most information of the original evaluation index system was obtained. The evaluation index system in the literature is re-simplified. The results show that the restructured simplified evaluation index system can retain more information compared with that in the literature, which proves the effectiveness of the method