Data-driven simulation analysis of different vehicle dispatching rules in e-hailing services

E-haling is getting more and more popular all around the world in the recent decade. It is unclear how different vehicle dispatching rules affect drivers’ income and satisfy passengers’ ride requests in E-hailing services. Based on the GPS trajectory and trip data of more than 1,000 taxis from Chengdu city, this paper simulates 3 different vehicle dispatching rules in e-hailing situation, and compares their performances in terms of 3 indicators, including the mean and difference of hourly incomes of all drivers, and satisfaction rate of service requests. The comparison experiments show that (1) comparing with the results based on real street-hailing data, the difference in drivers' income are significantly reduced under 3 dispatching rules in e-haling service; (2) the rule of high-valued dispatching leads to the lowest difference in drivers’ income and lowest average hourly income, (3) the rule of low-congested dispatching leads to the highest average hourly income and highest difference, and (4) the rule of nearest dispatching leads to a more balanced performance in all indicators.