EV Charging Forecasting for Grid Demand Response
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle-to-grid technologies lack effective forecasting methods for electric vehicle arrival and departure times, making it difficult to schedule charging and discharging for efficient energy storage and grid demand response.
Innovation Solution
A computer-based method forecasts total electric vehicle state of charge and energy storage capacity in parking areas by predicting future occupation, charge, and discharge capabilities, with adjustments based on actual arrivals and departures, using data from reservation systems and ANPR, to enable efficient energy storage and grid management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If reactive arrangements based on instantaneous state of charge are used, then the system responds to current conditions, but forward planning and collective demand management are ineffective
Solution Approach 1:
The system performs preliminary forecasting of EV arrivals, departures, and state of charge before actual events occur. This enables advance scheduling of charging and discharging operations, allowing the parking area to proactively manage energy resources rather than reacting to instantaneous conditions. The forecast is continuously updated with actual data to improve accuracy over time.
2Productivity
If forecasting methods are implemented, then forward planning is enabled, but the complexity of predicting unpredictable vehicle arrivals and departures increases
Solution Approach 1:
The system implements a practical forecasting approach that uses available data (arrival times, stay durations, charge/discharge capabilities) to generate useful predictions without attempting to model every possible variable. The forecast focuses on aggregate parking area state rather than individual vehicle predictions, reducing complexity while maintaining sufficient accuracy for demand response scheduling.
Solution Approach 2:
The forecasting system continuously updates its predictions by comparing forecasted values with actual measured data from arriving and departing vehicles. This feedback mechanism allows the system to learn from prediction errors and improve forecast accuracy over time, making the complex forecasting process self-correcting and more reliable.
3Productivity
If collective demand management is implemented, then efficient use of EVs for grid services is enabled, but the difficulty of coordinating multiple vehicles with varying capabilities increases
Solution Approach 1:
The system merges the capabilities of multiple individual EVs into a single aggregate forecast of total state of charge and total energy storage capacity for the parking area. This aggregation approach simplifies coordination by treating the parking area as a unified energy resource rather than managing each vehicle separately, while still accounting for variations in individual vehicle capabilities through the forecasting model.
Data Source
AI summary
A computer-based method is provided of forecasting total electric vehicle state of charge and total electric vehicle energy storage capacity in a parking area having parking bays with respective electrical vehicle charging and discharging facilities. Electric vehicles occupying the bays can provide temporary energy storage capacity to an energy grid. The forecasting method includes: forecasting future occupation of the bays by electric vehicles based on expected arrival times and expected stay durations; obtaining expected charge and discharge capabilities of the vehicles; forecasting a total state of charge and total energy storage capacity in the parking area; recording actual arrivals and departures; and adjusting the forecast of future occupation and the forecast of total state of charge and total energy storage capacity in the parking area.