A heterogeneous electric taxi fleet routing problem with recharging stations to maximize the company’s profit

¹ School of Industrial, College of Engineering, University of Tehran, Tehran, Iran
² School of Industrial Engineering, K. N. Toosi University of Technology, Tehran, Iran
³ Department of Industrial Engineering and Management Systems, Amirkabir University of Technology, Tehran, Iran

^* Corresponding author: mrabani@ut.ac.ir

Received: 28 July 2022
Accepted: 24 January 2023

Abstract

During the past years, many kinds of research have been done in order to reduce the cost of transportation by using different models of the vehicle routing problem. The increase in the amount of pollution caused by vehicles and environmental concerns about the emission of greenhouse gases has led to the use of green vehicles such as electric vehicles in the urban transport fleet. The main challenge in using electric vehicles with limited battery capacity is their long recharging time. For this purpose, several recharging stations are considered in the transportation network so that if the battery needs to be recharged, the electric vehicle can recharge and complete its journey. On the other hand, due to the limited amount of the electric vehicle’s energy, the fuel consumption of this fleet is highly dependent on their load, and it is necessary to consider their load in the planning. In this article, the problem of routing electric taxis is presented considering the economic and environmental aspects of implementing electric taxis for city services. Despite other studies that have only focused on reducing energy consumption or minimizing distance traveled by electric vehicles, for the first time, the problem of urban electric taxi routing has been modeled by considering different types of electric taxis with the aim of achieving the maximum profit of this business. The use of a heterogeneous fleet in this study leads to wider coverage of different types of demand. Therefore, a mathematical programming model is presented to formulate the problem. Then, several problem examples are designed and solved for validation purposes, and the simulated annealing algorithm (SA) will be introduced and used to solve large-scale problems.