【摘要】 为综合准确评估空管航班正常管理绩效,需对指标进行标定,降低其间关联性并保留有效信息.基于已有研究构建指标体系,针对现有数据特点采用灰色关联聚类与主成分分析法,对指标进行聚类分析及贡献率计算,依据累计贡献率期望、聚类数目等参数,提出用以表征所重构指标体系距理想状态相对距离的测度函数,为算法中临界值确定提供定量依据.以此得到最优聚类方案,实现指标标定与二级指标体系重构,避免了以往研究标定指标后对空管关键绩效领域反映不足的缺陷.并通过实证分析验证了测度函数的合理性.更多还原

【Abstract】 In order to estimate the regular flight management performance of air traffic management properly and comprehensively, a calibration of indices was required to reduce the relation and retain valid information. Based on the existing research, the index system was constructed. According to the characteristics of the existing data, the grey relational clustering and principal component analysis were used to cluster the indexes and calculate the contribution rate. According to the cumulative contribution rate expectation, the number of clusters and other parameters, a measure function was proposed to characterize the relative distance between the reconstructed index system and the ideal state, which provides a quantitative basis for determining the critical value in the algorithm. Based on the above method, the optimal clustering scheme was obtained, and the index calibration and secondary index system reconstruction were realized, thus avoiding the defect of insufficient reflection on key performance areas of air traffic control after previous research on calibration index. The rationality of the measure function was verified by empirical analysis.更多还原