The aim of this research is to study the optimal deployment of wireless charging stations (WCS) in urban transportation networks. It is widely acknowledged that the relatively short driving range of EV and the long battery charging times collectively lead to a phenomenon known as "range anxiety" of EV drivers. This phenomenon remains to be the major factor that hampers EV adoption. Thus, in this dissertation, we study a cost-effective deployment plan of WCSs that facilitates EV adoption by alleviating the two major causes of the “range anxiety” phenomenon.

In the first part of this dissertation, we propose a deployment plan that, for societal benefits, satisfies the charging demands of all EVs in the traffic network at the minimum investment cost. For this purpose, we formulate a new mathematical model to strategically deployWCSs in the traffic network in such a way that EVs can reach their destination without running out of energy. To solve the proposed model, we devise a combinedcombinatorialclassicalBenders Decomposition approach and enhance its efficiency further via employing surrogate constraints and an upper bound heuristic. The model and algorithm are tested on a real network with data from Chicago, IL for a sensitivity analysis and a deeper understanding of different design components of the wireless charging system.

In the second part, we illustrate that the WCS deployment plan can be greatly influenced by the frequently-changing traffic pattern in the road network under study. We demonstrate how a WCS network design, that is obtained based on input data of a single traffic period, might not be able to satisfy the charging demands during other traffic periods. We further show that even a WCS network design that is based on the peak traffic period might fail to satisfy the demands during less congested periods. That is, the peak traffic period is not the sole determinant of the optimal design. To that end, we study a robust deployment plan that is feasible and cost-effective across different realizations of traffic data. We build on the first part of this dissertation to propose a robust model where we consider the dynamic nature of the daily traffic patterns when we optimize the network design of the wireless charging infrastructure. We devise a customized Benders Decomposition approach to solve the proposed robust model, and we test the model and the algorithm on a real network data from Dallas, TX.

Finally, in the third part, we propose a new framework to plan the deployment of WCSs with the objective of influencing the routing behavior of EV drivers in an effort to improve the traffic assignment in the road network and alleviate congestion. For this purpose, we propose a new optimization model and test the applicability of the suggested approach on the famous Braess network and on Nguyen-Dupuis network. We illustrate, via the two examples, how an optimal WCS deployment can reform the traffic assignment and shift it from the (selfish-optimal) user equilibrium (UE) to the system (socially) optimal (SO) assignment. We further conduct sensitivity analyses to form a deeper understanding of the effectiveness of the suggested approach. The sensitivity analyses provide insights into the dependency of the traffic assignment on the EV population in the network, and on the attractiveness of the deployed WCSs for EVs.

Degree Date

Fall 2018

Document Type



Engineering Management, Information, and Systems


Halit Uster



Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Available for download on Thursday, August 27, 2020