Mobile EV Charger Deployment Using Multi-Agent Reinforcement Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The expansion of fixed electric vehicle charging facilities cannot keep pace with the increasing demand for electric vehicle charging, and there is a need for efficient distribution and operation of mobile electric vehicle chargers to address this demand.
Innovation Solution
A multi-agent reinforcement learning-based method for optimizing the deployment of mobile electric vehicle charging stations using a simulation system and reinforcement learning technology to maximize charging profit and minimize movement costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If mobile charging stations are deployed to high-demand regions, then charging service coverage and profitability are improved, but system complexity and operational coordination difficulty increase
Solution Approach 1:
The system divides the charging service area into multiple grid zones and assigns mobile charging stations to specific zones. Each station operates independently within its assigned grid, making decisions based on local charging demands rather than coordinating complexly with all other stations. This segmentation reduces system coordination complexity while maintaining comprehensive charging coverage across the entire service area.
Solution Approach 2:
The patent implements dynamic zone assignment where mobile charging stations can transition between different grid zones based on real-time charging demands. The system dynamically adjusts station locations and zone boundaries to match demand patterns, allowing the system to scale and adapt without increasing coordination complexity proportionally.
2Adaptability or versatility
If multiple mobile charging stations are deployed, then charging demand coverage is improved, but movement costs and operational complexity increase
Solution Approach 1:
The service area is divided into discrete grid zones, and each mobile charging station is assigned to serve specific zones. This segmentation allows the system to deploy the minimum necessary number of stations to cover high-demand areas, avoiding redundant deployment and unnecessary movement costs while maintaining adequate charging demand coverage.
Solution Approach 2:
The system uses historical charging demand data and AI prediction models to pre-identify high-demand zones and pre-position mobile charging stations in these areas before peak demand occurs. This preliminary action reduces the need for frequent long-distance movements, lowering movement costs while ensuring charging demand coverage is maintained.
3Productivity
If AI prediction and multi-agent reinforcement learning are used for station deployment, then charging profit is improved, but computational requirements and system complexity increase
Solution Approach 1:
The AI prediction and reinforcement learning processes are segmented and distributed across multiple independent agents, each responsible for a specific grid zone. Each agent independently learns and makes deployment decisions for its zone based on local demand patterns, avoiding the need for a single complex centralized computation system while maximizing charging profit through localized optimization.
Solution Approach 2:
The system employs self-learning reinforcement learning agents that automatically improve their deployment strategies through continuous interaction with the environment and receipt of reward signals. These agents autonomously optimize charging profit without requiring external retraining or complex manual configuration, reducing system computational complexity over time as the agents become more proficient.
Data Source
AI summary
A multi-agent reinforcement learning-based mobile electric vehicle charging service method may include: generating an electric vehicle charging demand; detecting a state of a mobile charging station; and determining an action of the mobile charging station including moving or waiting based on the electric vehicle charging demand and the state of the mobile charging station. The method may further include: paying a reward as a feedback with respect to a result including a charging profit and a moving cost based on the determined action; storing and accumulating the action, the result, and the reward as learning data; and training a multi-agent reinforcement learning model for generating the optimal deployment of the mobile charging station by using the accumulated learning data.