Master Thesis Defense - March 17, 2015: Demand Forecasting and Location Optimization of Recharging Stations for Electric Vehicles in Carsharing Industries

Elnaz Moein Namin

Tuesday, March 17, 2015 at 2:00
Room EV003.309

You are invited to attend the following M.A.Sc. (Quality Systems Engineering) thesis examination.

Examining Committee

Dr. J. Clark, Chair
Dr. A. Awasthi, Supervisor
Dr. J. Bentahar, CIISE Examiner
Dr. B. Jaumard, External Examiner (CSE)

Abstract

Carsharing is an alternative to private car usage. Using electric-vehicles as a substitute to fuel vehicles is a wiser option which leads to lower fuel emissions, more energy savings and decreased oil dependency. However there are some barriers in using electric vehicles at large scale in carsharing companies. Battery power limitation and lack of sufficient infrastructure are some of them. Accurate demand forecasting is a must for this purpose.

In the first part of this thesis, we investigate the demand forecasting problem for carsharing industries and apply four techniques namely simple linear regression, seasonally adjusted forecast, Winter's Model and artificial neural networks to decide the right number of vehicles to be made available at each station to meet the customer requests. The results on randomly generated test datasets show that artificial neural networks perform better over the other three.

In the second part, we investigate the location planning problem of recharging stations for electric vehicles. The base model used for this study is the mathematical optimization model proposed by Wang & Lin (2013). Firstly, we improve their MIP model and solve it using AIMMS (Advanced Interactive Multidimensional Modeling System). Secondly, we propose Genetic Algorithm for the same problem and implement it in Matlab. The obtained results are compared with previous work done by Wang and Lin (2013). The comparisons show better performance of the proposed methods.