Bike Sharing Data Analytics for Smart Traffic Management

With the rapid development of the sharing economy, bike sharing has become an indispensable part of urban transportation because of its zero emissions and low energy consumption. The Quality of live (QoL) of urban citizens can benefit from a well-maintained bike sharing system with reasonable distributions. Whereas, rebalancing the bike sharing system in large-scale urban areas is technically challenging, considering the limited docks, dynamic user demands, as well as finite trucks with limited capacity during the rebalancing work. Initially, we propose a cluster algorithm to filter noises and accommodate complex terrain with the distance and flow between stations during station clustering. After clustering, the global rebalancing problem can be transmitted into the sub problem as inner-cluster truck path routing for each single cluster, and a corresponding routine algorithm is presented. Finally, the experimental results with real-world datasets prove the efficiency and superiority of our proposed solutions.