Optimization of bus lines based on passenger group moving behaviors

Optimization of public bus routes in a transportation system is one of the keys to achieve smart city concept, delivers better quality, and cost-saving travel to the passengers. Compared to private vehicles, public transportation system has been chosen by millions of passengers each day with different routes, generates certain travel patterns. From these travel patterns, routes are optimized by the transportation experts to reduce the gap between existing routes and travel demands. Thus, the transportation system is depending upon experts to optimize it manually. This paper focuses on discovering the mobility pattern of passengers with buses and subways to optimize the routes of Beijing. To this end, we first illustrate the changes in certain bus lines by the experts in a specific interval of time during a year and their impact. Based upon mobility features, we propose Group Travel Demand Identification (GTDI) method that suggests the route adjustments for optimization of bus lines. We also demonstrate a comparison of our method with other optimization techniques. Route forecasting results are matched with the adjustments of experts to validate the performance and reliability, To solve route adjustment problem, we conducted an extensive study on sets of data records and adjustment records of bus lines collected from Beijing Public Transport Corporation (BPTC). To generate mobility pattern from Smart Card Data, we studied more than 845 million passengers.