2-III_am

[113] Optimization model to allocate taxis to railway stations focusing on increasing demand during rainfall

Linear programming formulation and its application to taxi probe data in Tokyo

○Keitaro Morimoto¹, Kazuki Tanno¹, Kengo Hamada, Koki Ogai, Ken-ichi Tanaka² (1. Graduate School of Science and Technology, Keio University, 2. Faculty of Science and Technology, Keio University)

Keywords:taxi allocation model, linear programming model, taxi probe data, machine learning

We propose a mathematical optimization problem that allocates empty taxis to railway stations so as to capture taxi demand. Input demand data is estimated by machine learning using taxi probe data in Tokyo area. Two measures are employed in the mathematical optimization model. One is the sum of the total demand captured at stations and on the way to the station. The other is the total travel distance for all taxis from the original location to the location at which demand is captured. By simplifying some assumptions, the proposed model is formulated as a linear programming problem. We applied the proposed model to taxi probe data in Tokyo, and obtained solutions that well balance the captured demand and the total travel distance for taxi allocation.