Subsidy Model for Taxi Drivers in Beijing


This research was completed as a project in China Undergraduate Mathematical Contest in Modeling and win the First Prize. Research data was acquired from AutoNavi corporation.


In China, especially in metropolis like Beijing, there exists huge difficulty in taking taxi. In order to quantify the difficulty and find the optimal solution, two models were built in this paper. Data provided by AutoNavi corporation include the exact position of each taxi and whether there is a customer on or not. In the first model, probability distribution function (pdf) of waiting time was derived based on the operational data from AutoNavi. Expected waiting time was employed as the “difficulty index” and can be calculated from pdf easily. The optimal subsidy model can be developed on the basis of the first difficulty quantifying model. Optimization was realized by the iteration of simulation and adjustment of model parameters.


The paper in this competition can be downloaded [here] ( in Chinese)