Electric Vehicle Charging Station Resource Allocation: A Data-Driven Robust Optimization Approach

• Author / Creator

  Rouhani, Mohammadhadi

• The adoption of electric vehicles has been growing steadily in recent years, and projections indicate that this trend will continue. However, the availability and capacity of charging stations have not kept pace with this growth, leading to long wait times and congestion at charging stations. The installation and operation of electric vehicle charging stations (EVCSs) are non-trivial problems and require careful consideration of several factors, including the size of the charging station. This involves determining the optimal number of charging units and their capacity to meet the expected charging demand. This becomes even more complex when the charging station is coupled with an on-site solar photovoltaic (PV) panel system and a battery energy storage system (BESS).

  The goal of sizing EVCSs is to create a methodology that can enhance the utilization of charging infrastructure, decrease waiting times, and enhance the user experience. An efficient sizing strategy will guarantee that the charging stations can meet the projected demand while keeping installation and operating expenses to a minimum. Moreover, the proper implementation of charging infrastructure is crucial to the widespread adoption of EVs, as the availability of charging infrastructure is a key factor in consumer decision-making.

  We study two problems in this thesis:

  Our first task involves a two-stage sizing assessment of a single EVCS that is co-located with on-site PV and BESS systems. Initially, we want to identify EVCS sizing options that meet a blocking rate threshold, which is a user experience performance metric. Subsequently, for each optimal-sized EVCS option, we recommend robust sizing solutions for the PV and BESS systems to minimize the reliance on the main power grid. We address this problem using convex optimization and introduce a Chebyshev inequality for robust sizing. Our simulation results establish a correlation between the sizing of the PV and BESS systems and confirm that larger sizing of PV and BESS is required for reduced dependency on the main grid. Additionally, we discover that a significant PV system is necessary for an EVCS to rely entirely on solar energy without the assistance of a BESS. Thus, we recommend combining a PV system with a BESS for optimal performance.

  In the second problem, we proceed to assess a network of EVCSs and establish an optimization problem to optimally size each EVCS in the network, subject to various constraints. These constraints stem from performance metrics (such as response time) and total costs, encompassing both capital and operating costs, across all locations. Due to the complex nature of this queueing system optimization problem, we must make certain assumptions to obtain a feasible solution. We solve the optimization problem for a small-scale network of EVCSs including the traffic flow. Our findings indicate that when optimizing a network of EVCSs, the sizing alternative at each location is likely to have a direct impact on the performance metrics.

• Subjects / Keywords

  • Data driven
  • Electric vehicle
  • Charging station
  • Robust optimization
  • Resource allocation

• Graduation date

  Fall 2023

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-cf9k-bc96

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.