Optimal lane expansion model for a battery electric vehicle transportation network considering range anxiety and demand uncertainty

Abstract

This paper investigates a novel design problem involving an optimal government lane expansion scheme for a battery electric vehicle (BEV) transportation network. A lane expansion model is established considering the BEV charging time, driver range anxiety and uncertain transportation demand. This model aims to minimize the total travel time (i.e., sum of the driving time and charging time) of all drivers in the transportation network and optimize the lane expansion scheme (i.e., the number and location of extended lanes in the network) under the established investment ceiling. To address demand uncertainty, an improved adjustable robust optimization method is further proposed to relax the model by introducing two control parameters. Based on the framework of the active set algorithm, a local optimal solution algorithm is designed to effectively solve the abovementioned model. Column generation is embedded in the above algorithm to avoid path enumeration. Sensitivity analyses are conducted for different control parameters and government investment scales. The results show that the model and algorithm we proposed can provide a theoretical basis for the government to improve the traffic efficiency of the transportation network and achieve the goal of sustainable transport.