An Adaptive Large Neighborhood Search Approach for Electric Vehicle Routing with Load-Dependent Energy Consumption

Abstract

Electric vehicles are gaining popularity day-by-day aided by growing pollution concerns with fossil fuel vehicles. Many logistics companies have already started testing electric vehicles for deliveries in cities. However, electric vehicles have issues such as range anxiety and long recharge times. These issues have to be considered in routing electric vehicles to avoid inefficient routes. One of the important factors that affects the amount of battery consumed is load carried by the vehicle. Considering loads will significantly affect the routes determined in the electric vehicle routing problem (EVRP). Most previous studies solved EVRP with distance minimization as the objective. We have considered load of vehicle in the power estimation function to calculate the energy requirement. An adaptive large neighborhood search (ALNS) with special operators particular to this problem structure is presented. ALNS was tested on 56 benchmark instances and it found better solutions for 14 instances and for 15 instances the solutions matched the best-known solution.