EV Charging Matchmaking Using SoC Maps and Reward Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for scheduling electric vehicle charging sessions are suboptimal in balancing energy supply and demand over a wider geographical area, leading to uncertainty in charging station availability and scheduling for electric vehicle drivers.
Innovation Solution
A cloud-based system that uses dynamic bidirectional matching of electric vehicle charging demand with available charging station supply within a geofenced area, employing state-of-charge maps, power maps, and reinforcement learning to optimize energy distribution and scheduling, including features like adaptive routing, load balancing, and dynamic pricing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If one-to-one charging session scheduling is used, then individual EV charging needs are met, but energy supply and demand balance over wider geographical area deteriorates
Solution Approach 1:
The patent combines multiple EV charging demands with multiple EVSE supplies into a unified many-to-many matching system. The cloud-based platform aggregates charging requests from multiple EVs and available capacity from multiple charging stations, then performs optimized matching that balances energy supply and demand across the entire geofenced region rather than handling individual one-to-one pairings.
Solution Approach 2:
The cloud-based server performs multiple functions simultaneously: it collects EV information and charging station information, generates SoC maps and power maps, predicts future energy supply and demand, performs many-to-many matching optimization, and provides routing recommendations. This multi-functional approach enables both individual charging needs and regional balance to be addressed.
2Adaptability or versatility
If dynamic bidirectional matching is implemented, then energy supply and demand balance improves, but system complexity increases
Solution Approach 1:
The cloud-based server acts as an intermediary between EVs and charging stations. It receives information from both sides, performs complex processing including map generation, prediction, and optimization algorithms, then returns simplified matching recommendations to EVs and scheduling information to charging stations. This intermediary approach manages system complexity by centralizing computational burden.
Solution Approach 2:
The system performs preliminary actions by generating SoC maps and power maps, and predicting future energy supply and demand before actual charging sessions occur. These pre-computed data structures and predictions enable efficient real-time matching without requiring complex calculations during the matching moment itself.
3Reliability
If cloud-based dynamic prediction is used, then charging station availability accuracy improves, but data processing requirements increase
Solution Approach 1:
The cloud-based server performs preliminary data processing by generating SoC maps from EV information and power maps from charging station information, and by predicting future energy supply and demand trends. These pre-computed maps and predictions organize raw data into structured formats that enable accurate availability forecasting without requiring intensive real-time processing for each charging request.
Solution Approach 2:
The system creates simplified representations (maps) of the complex charging ecosystem - SoC maps represent the state of charge distribution across EVs, and power maps represent available charging capacity at stations. These map copies enable efficient querying and prediction without processing all underlying raw data for each operation.
Data Source
AI summary
A method for dynamically matching energy demand of a population of electric vehicles (EVs) with energy supply of a population of charging stations within a geofenced perimeter includes receiving, via a cloud-based server, EV information from each respective EV, and receiving charging station information from each respective charging station. The method includes generating an SoC map and a charging station power map from the EV information and the charging station information, respectively, and predicting the energy supply and demand using the maps. The server dynamically matches the EVs to at least one of the charging stations or vice versa using a reward function, the predicted energy supply, and the predicted energy demand, including generating a rank-ordered listing for each of the EVs and/or each of the charging stations in a manner that maximizes an expected discounted future reward.