azra ghobadi

An effective approach to tackling climate change is to replace electric vehicles with their diesel counterparts. Although this replacement has many challenges, it is not impossible. In this paper, two multi-objective mathematical models - multiple deposit problem routing - charging station location and battery replacement of heterogeneous freight electric vehicles were presented. In addition, to increase customers' satisfaction, fines for violating time windows were applied in the models. Each of these two models pursues three goals. The first goal is to minimize (the total cost of routing, the cost of constructing charging stations or replacing the battery, and the cost of breaking time windows); the second goal is to minimize the number of used cars and the third goal is to minimize the number of charging or battery replacement stations. The models were solved by CPLEX solver with GAMS software in small size. Examination of the results shows that in the first proposed model, battery replacement stations were used. Fewer cars were used. Also, in the second proposed model that the stations offer partial charging; the total cost was reduced. Also, in both models, the cost decreases with increasing the number of depots.

The use of an Electric Vehicle (EV), particularly in different operations of goods distribution is a solution for salvaging the crowded cities of the world from air and noise pollutions as well as Green House Gas (GHG) emission. This paper presents a Multi-Depot Electric Vehicle Routing Problem (MD-EVRP) with recharging stations by considering the expected penalty of fuzzy time windows in pickup/delivery. Since the MD-EVRP with Fuzzy Time Windows and Pickup/Delivery (MD-EVRP-FTW-PD) constraints is an NP-hard problem, three meta-heuristics (i.e., Simulated Annealing (SA), Variable Neighborhood Search (VNS) and a hybrid of SA and VNS (VNS-SA)) are used to solve such a hard problem. The parameters of these algorithms are measured by the Taguchi experimental design method. The proposed hybrid VNS-SA algorithm is more efficient in comparison with other algorithms.