Real-Time EV Charging Scheduling With Demand Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electric Vehicle Charging Stations (EVCS) face challenges in accurately predicting and optimizing EV charging demand due to uncertainties in EV arrival times, individual power requirements, and varying driving patterns, leading to mismatches between day-ahead power procurement and real-time scheduling, resulting in increased costs.
Innovation Solution
A system for real-time scheduling EVs at EVCS, utilizing a processor-based system that forecasts demand using ARIMA techniques, optimizes demand profiles through linear programming, and schedules charging by adjusting parking times and enabling peer-to-peer energy transfers to minimize costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional EV charging scheduling methods are used, then charging demand can be predicted, but there is a mismatch between day-ahead power procurement and real-time scheduling leading to increased costs
Solution Approach 1:
The system performs preliminary actions by forecasting EV arrival patterns and charging demand in advance, optimizing the demand profile before real-time scheduling. This allows the EVCS to procure power more accurately in day-ahead markets, reducing the mismatch between procured and actual power needs, thereby lowering procurement costs while maintaining reliable demand prediction
Solution Approach 2:
The system dynamically adjusts the charging schedule in real-time based on actual EV arrivals and departures, rather than sticking to a static day-ahead plan. This dynamic scheduling allows the EVCS to adapt to uncertainties in EV behavior patterns, improving both prediction reliability and cost efficiency by optimizing power usage as conditions evolve
2Productivity
If EVCS procures power based on forecasted demand, then power can be secured in advance, but uncertainties in EV arrival times and power requirements lead to scheduling mismatches
Solution Approach 1:
The system performs preliminary optimization of the demand profile based on forecasted EV arrivals and charging patterns before real-time execution. This preliminary action enables efficient day-ahead power procurement while the optimized profile serves as a flexible guide rather than a rigid plan, allowing adaptations when actual EV behavior deviates from forecasts
Solution Approach 2:
The system implements real-time monitoring and feedback mechanisms that compare actual EV arrivals and charging demands against the forecasted profile. This feedback enables dynamic adjustments to the scheduling in real-time, maintaining adaptability to uncertainties while preserving the efficiency benefits of advance power procurement through the optimized baseline profile
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
This disclosure relates generally to real-time scheduling EVs for charging at electric vehicle charging station (EVCS). The EVCS enables charging of EVs, procure power from open markets at time varying prices. The estimation of demand profile for EVCS is very challenging as arrival of Electric Vehicles (EVs) is not known to EVCS apriori and is also heavily dependent on several uncertain factors associated with the EVs such as EVs individual power requirements, driving and travel patterns, EV owner behavior etc. The disclosure estimates a demand profile of an EVCS, optimizes the demand profile and schedules EV arrivals in real time. The scheduling is performed by offering flexibility by 1) increasing parking hours so that charging sessions can be scheduled in lower price periods and 2) participating in peer-to-peer transfer, so that EV with extra parking time charged during lower price periods can transfer power to others during high price periods.