Hierarchical framework for demand prediction and iterative optimization of EV charging network infrastructure under uncertainty with cost and quality-of-service consideration

Abstract

As electric vehicles (EVs) continue to grow in popularity, the demand for efficient and effective EV charging infrastructure is rapidly increasing. However, the design and deployment of such infrastructure faces numerous challenges, some of which include demand uncertainty, EV network layout, maintenance cost, and quality of service (QoS), and have proven to be particularly challenging in cases where all objectives need to be addressed simultaneously. To address these challenges, this study proposes a hierarchical iterative optimization approach that aims to minimize the demand–supply mismatch at a higher level and considers both cost and QoS under uncertainty at a lower level in a dynamic EV charging network. Specifically, the proposed approach uses a time-series linear regression ensemble to accurately predict future demand at demand points, which is then used to match demand at demand points with supply at supply points. We then model the infrastructure allocation problem as a knapsack problem and use an allocation algorithm to determine the optimal number of charging stations at each supply point. The infrastructure is allocated to minimize total cost and maximize QoS. A framework for EV infrastructure optimization is designed and can be applied to any type of network. The effectiveness of our proposed approach is demonstrated through extensive experimentation on various datasets by comparing with existing state-of-the-art approaches. Simulation results show that the proposed approach is robust to different EV networks, handles uncertainty well, can effectively address the challenges associated with EV infrastructure optimization, and can provide policy makers and infrastructure planners with a practical and efficient tool for designing and deploying charging infrastructure. This study contributes to the field of EV infrastructure network optimization by modeling QoS and proposing a hierarchical optimization framework that effectively addresses the challenges associated with the design and deployment of charging infrastructure.