EV Charging Forecasting for Multi-Vehicle Site Power Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accurate forecasting of electric vehicle (EV) charging and discharging at a site is difficult due to stochastic arrival patterns, especially in facilities with multiple EVs, leading to challenges in power management and financial burdens.
Innovation Solution
A computer-implemented method using machine learning (ML) models trained on historical and real-time data to predict EV charging and discharging, incorporating data from various sources like EVs, traffic, weather, and user schedules, to optimize energy management and trading.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are trained on historical and real-time data to predict EV charging and discharging, then prediction accuracy is improved, but system complexity increases
Solution Approach 1:
The system divides the prediction task into multiple components: data collection from various sources (EVs, traffic, weather, user schedules), data processing, model training, and inference. The ML models are segmented into different types (charging prediction models, discharging prediction models) that can be trained independently on specific historical data patterns, reducing the complexity of training a single monolithic model while maintaining high prediction accuracy.
Solution Approach 2:
The patent introduces an intermediary data processing layer that collects, cleans, and prepares data from multiple sources before feeding it to the ML models. This intermediary layer includes data normalization, feature extraction, and temporal aggregation functions that simplify the interaction between raw data and models, reducing system complexity while improving prediction accuracy.
2Adaptability or versatility
If multiple EVs are served at a facility with limited power capacity, then service versatility is improved, but power management difficulty increases
Solution Approach 1:
The system performs preliminary actions by predicting EV arrivals, charging needs, and discharging capabilities in advance using historical data and real-time inputs. The ML models forecast charging demand and available discharging power before the actual charging events occur, enabling the power management system to pre-allocate resources and plan charging schedules, thereby reducing real-time decision complexity while serving multiple EVs.
Solution Approach 2:
The power management system dynamically adjusts charging and discharging operations based on real-time predictions and actual conditions. The system continuously updates its power allocation strategy based on predicted EV arrivals, battery states, and grid conditions, allowing flexible adaptation to varying service demands while maintaining manageable complexity through automated dynamic control.
Data Source
AI summary
A computer-implemented method of predicting electric vehicle (EV) charging and discharging. The method includes: collecting input data from a database and a plurality of EVs authorized to be charged or discharged at a site, the input data including historical input data and real time input data associated with charging and discharging the EVs; training a plurality of machine learning (ML) models using the historical input data to predict EV charging and discharging outcomes at the site for a time period; and predicting, by an ML inference device that is applying the trained ML models to the real time input data, an amount of power needed for EV charging or an amount of EV discharging power available by the one or more authorized EVs during a time interval.