A dynamic pricing method for carpooling service based on coalitional game analysis

In recent years, carpooling service provided by corporations like Uber (UberPool), Didi (DidiPool) and Lyft (Lift Link) have become more and more popular. It helps alleviating the urban traffic congestion, by decreasing the empty seats rate. To balance the supply and demand of the taxi service, a dynamic pricing method is needed. More specifically, passengers taking a same vehicle may be charged differently, even thought they shared a most part of a trip. It often challenges the current dynamic pricing policy that how to balance the service and the pricing among different passengers who shared a certain route in their personal trip. In view of this challenge, we propose a new dynamic pricing method and divide the payoff according to the contribution of each passenger. Concretely, we deploy the framework of coalitional game to analyze spatial temporal constraints that guarantee individual benefits from the carpooling coalition. Then, we explore the Nash Product to maximize the utility of passengers as a whole and reduce our problem into a geometry-programming problem. At last we use Shapley value method to measure the specific contribution of each passenger. We conduct a simulated experiment and the results show effectiveness of our method.