Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Resilient Charging Infrastructure via Decentralized Coordination of Electric Vehicles at Scale

Chuhao Qin, Alexandru Sorici, Andrei Olaru, Evangelos Pournaras, Adina Magda Florea

TL;DR

This work addresses the challenge of decentralized EV charging at scale under uncertainty, proposing DECharge, a framework that balances driver comfort and system efficiency through adaptive charging behaviors and collective learning. It combines I-EPOS-based decentralized coordination with a behavior-recommendation mechanism to navigate Pareto-optimal trade-offs across spatio-temporal demand and station capacities, while preserving privacy. Empirical results on real-world Paris charging data and temporally distributed demand show significant reductions in waiting times and more balanced station load, with strong resilience to outages and adversarial behavior. The approach offers scalable, privacy-preserving coordination for resilient EV charging infrastructure and provides open-source baselines for future research.

Abstract

The rapid adoption of electric vehicles (EVs) introduces major challenges for decentralized charging control. Existing decentralized approaches efficiently coordinate a large number of EVs to select charging stations while reducing energy costs, preventing power peak and preserving driver privacy. However, they often struggle under severe contingencies, such as station outages or unexpected surges in charging requests. These situations create competition for limited charging slots, resulting in long queues and reduced driver comfort. To address these limitations, we propose a novel collective learning-based coordination framework that allows EVs to balance individual comfort on their selections against system-wide efficiency, i.e., the overall queues across all stations. In the framework, EVs are recommended for adaptive charging behaviors that shift priority between comfort and efficiency, achieving Pareto-optimal trade-offs under varying station capacities and dynamic spatio-temporal EV distribution. Experiments using real-world data from EVs and charging stations show that the proposed approach outperforms baseline methods, significantly reducing travel and queuing time. The results reveal that, under uncertain charging conditions, EV drivers that behave selfishly or altruistically at the right moments achieve shorter waiting time than those maintaining moderate behavior throughout. Our findings under high fractions of station outages and adversarial EVs further demonstrate improved resilience and trustworthiness of decentralized EV charging infrastructure.

Resilient Charging Infrastructure via Decentralized Coordination of Electric Vehicles at Scale

TL;DR

This work addresses the challenge of decentralized EV charging at scale under uncertainty, proposing DECharge, a framework that balances driver comfort and system efficiency through adaptive charging behaviors and collective learning. It combines I-EPOS-based decentralized coordination with a behavior-recommendation mechanism to navigate Pareto-optimal trade-offs across spatio-temporal demand and station capacities, while preserving privacy. Empirical results on real-world Paris charging data and temporally distributed demand show significant reductions in waiting times and more balanced station load, with strong resilience to outages and adversarial behavior. The approach offers scalable, privacy-preserving coordination for resilient EV charging infrastructure and provides open-source baselines for future research.

Abstract

The rapid adoption of electric vehicles (EVs) introduces major challenges for decentralized charging control. Existing decentralized approaches efficiently coordinate a large number of EVs to select charging stations while reducing energy costs, preventing power peak and preserving driver privacy. However, they often struggle under severe contingencies, such as station outages or unexpected surges in charging requests. These situations create competition for limited charging slots, resulting in long queues and reduced driver comfort. To address these limitations, we propose a novel collective learning-based coordination framework that allows EVs to balance individual comfort on their selections against system-wide efficiency, i.e., the overall queues across all stations. In the framework, EVs are recommended for adaptive charging behaviors that shift priority between comfort and efficiency, achieving Pareto-optimal trade-offs under varying station capacities and dynamic spatio-temporal EV distribution. Experiments using real-world data from EVs and charging stations show that the proposed approach outperforms baseline methods, significantly reducing travel and queuing time. The results reveal that, under uncertain charging conditions, EV drivers that behave selfishly or altruistically at the right moments achieve shorter waiting time than those maintaining moderate behavior throughout. Our findings under high fractions of station outages and adversarial EVs further demonstrate improved resilience and trustworthiness of decentralized EV charging infrastructure.

Paper Structure

This paper contains 20 sections, 17 equations, 12 figures, 4 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. System Model
  4. Definitions and assumptions
  5. Problem formulation
  6. Detailed System Design
  7. Charging option generation strategy
  8. Collective learning for optimized selection
  9. Charging behavior recommendation
  10. Experimental Methodology
  11. Dataset and implementation details
  12. Experimental scenarios and algorithms
  13. Evaluation metrics
  14. Baselines
  15. Performance Evaluation
  16. ...and 5 more sections

Figures (12)

  • Figure 1: An example of DECharge where an EV selects one of three charging stations. The EV observes the travel time ($T$) and estimated queuing time ($Q$) at each station, and chooses its behavior (selfish, moderate or altruistic) according to $T+Q$. After all EVs make their choices, the actual queuing time ($AQ$) at each station is determined via coordination. This coordination guides EVs toward moderate behavior, effectively reducing their actual waiting time.
  • Figure 2: A case where two EVs choose the same charging station without coordination. EV 1 requests charging time of 30 min and EV 2 requests 10 min. They start to request charging at 10:00, but reach the station at 10:00 and 10:10 respectively. The coordinate axis shows the update of actual queuing time at charging station $M$.
  • Figure 3: The overall system framework of DECharge for decentralized EV charging control ($Q$ is the queuing time (minute) observed by EVs; $T$ is the travel distance (km) from EVs to stations).
  • Figure 4: An example of the learning iteration of I-EPOS based on a binary tree topology with $7$ agents (EVs). During the bottom-up phase, each agent sends its charging option $k$ or the branch response $B$ to its parent. During the top-down phase, each agent sends the aggregated options $G$ and decision $\delta$ to its children.
  • Figure 5: Spatial distribution of charging stations in Paris and their frequency of "Charging" state (i.e., the number of times each charging station has been requested by EVs over a year).
  • ...and 7 more figures

Theorems & Definitions (3)

  • proof
  • proof
  • proof