Validity research of judging boarding stop by cluster analysis smart card data

Transaction data of Automatic Fare Collection in public transportation record travel time, travel route number of passengers, etc. Analyzing and mining these data, many travel characteristics and laws of transit passengers can be attained, which are important to improve the level of public transit planning and management. Judging boarding stop of passengers is a basis of smart card analysis. Judging boarding stop by cluster analysis of boarding time was presented by Southeast University. However, Up to now, the validity of the method hasn't researched. This paper collected smart card data and relevant data of Shenzhen City, and tested the method. Then the paper evaluated precision of the method through field investigation data of bus operation and analyzed the influence factors. Finally, the improved method was presented.