Intelligent Vehicle Charging Policy via Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optimizing the charging of multiple electric vehicles at a charging site is challenging due to varying arrival and departure times, states of charge, and the need for precise charge management to ensure vehicles are fully charged by their departure times, which existing technologies struggle to address effectively.
Innovation Solution
A vehicle charging system utilizing a Policy Gradient Algorithm and simulation modeling to develop an optimal charging policy, where a learning model aggregates data and simulates scenarios to determine the best strategy for charging multiple vehicles, incorporating factors like AC and DC power sources, renewable energy, and carbon emissions, without requiring extensive real-world data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a learning model with simulation modeling is used to optimize charging policy, then charging efficiency and cost minimization improve, but system complexity increases
Solution Approach 1:
The patent creates a simulated charging environment that copies real-world charging scenarios to train the learning model. This simulation model replicates vehicle arrivals, departures, charging rates, and infrastructure constraints without requiring extensive real-world data collection, thereby improving charging efficiency while managing system complexity through virtual modeling.
Solution Approach 2:
The learning model is pre-trained in the simulated environment before deployment to real charging sites. This preliminary training phase allows the system to learn optimal charging policies in advance through extensive simulation runs, so that when deployed, the model can make informed decisions without real-time complexity overhead.
2Measurement precision
If real-world data is used to train the learning model, then model accuracy improves, but data availability and quantity become limiting factors
Solution Approach 1:
Instead of relying on scarce real-world data, the patent uses simulation modeling to generate synthetic training data that replicates real charging scenarios. The simulation creates virtual vehicle arrivals, departures, states of charge, and infrastructure conditions, providing abundant training data without being limited by actual data collection constraints.
Solution Approach 2:
The simulation environment acts as an intermediary between the learning model and real-world charging operations. It provides a middle ground where the model can be trained with synthetic data that mimics real conditions, bridging the gap between limited real data and the need for comprehensive training without requiring extensive real-world data collection.
3Adaptability or versatility
If the system adapts to changes in charging infrastructure and vehicle conditions, then adaptability improves, but the need for user input and future state predictions increases
Solution Approach 1:
The learning model is designed to autonomously adapt to changes in charging infrastructure and vehicle conditions without requiring user input. The model continuously learns from simulation data and real-world feedback, automatically adjusting charging policies to new conditions such as different vehicle types, charging rates, and infrastructure constraints, making the system self-adapting and easy to operate.
Solution Approach 2:
The system incorporates feedback mechanisms where the learning model receives information about actual charging outcomes and uses this feedback to improve future decisions. By learning from both simulation results and real-world charging data, the model adapts to changing conditions automatically, reducing the need for manual user input while maintaining high adaptability.
Data Source
AI summary
Systems and methods for vehicle charging are disclosed. The system is configured to aggregate available data associated with states of multiple vehicles and a charging site and associated with charging the multiple vehicles at the charging site. The system is also configured to inference a pre-trained learning model to apply a charging policy to the available data to charge the vehicles at the charging site. The pre-trained learning model includes one or more learning agents configured to take actions and to observe effects of the actions in a simulated charging environment to obtain the charging policy.