Passenger flow prediction of subway systems utilizing TSK fuzzy modeling based on Gustafson-Kessel Possibilistic c-Means Clustering approach

In recent years many cities around the world are experiencing increasing demand in the transportation sector. The subway transportation system is an important public transit system component. Because of crowdedness in trains and insufficient capacity, transit managers need to predict passenger flows inadvance to consider different policies for the changes in passenger demand, reduce congestion, and maintain the acceptable service quality. In this paper, we propose a new TSK fuzzy logic system (GKPCM-TSK) to predict hourly passenger arrivals a week earlier. In our proposed model Gustafson-Kessel Possibilistic c-Means Clustering algorithm (GKPCM) is used to modify the rule extraction step in Tagaki–Sugeno–Kang (TSK) fuzzy system, in which the optimized number of clusters is determined using GKPCM. The proposed methodology is applied to 6 stations of line 3 of the Tehransubway system. The results indicate that the proposed methodology performs well in the hourly prediction of passenger arrivals.

کلیدواژه ها:

Passenger arrivals prediction ، TSK fuzzy logic system ، Public Transportation ، Subway systems.