Clustering Algorithm for Railway Container Handling Stations Based on Container Operation Data
To eliminate the impact of different container applications on the quality evaluation of container manufacturing,this paper analyzed the container operation and maintenance data from the past 10 years that covered multiple factors such as transportation categories,station characteristics,and handling frequency.Based on the differences in the operation of railway container stations,a container cluster index system was constructed using container operation data.A dataset available for cluster analysis was formed through data preprocessing technology.Various clustering algorithms,such as K-means,EM,and Canopy,were utilized to comprehensively analyze the processed dataset to identify station groups with similar container operation characteristics.To overcome the potential bias of different algorithms,an algorithm fusion strategy based on the voting principle was introduced to reduce the dependence among algorithms.The results show that the fused cluster effect is superior to the effect of any single algorithm,which can provide basic data for the comprehensive analysis of railway container maintenance.
RailwayIntegrated TransportationContainer MaintenanceData MiningAnalysis of Loading and Unloading StationCluster AlgorithmFusion Algorithm
国家科技期刊平台
NSTL
万方数据
