Pattern-Based Charging for EV Battery Health
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electrified vehicle batteries experience reduced capacity and performance due to prolonged high state of charge, which existing charging methods fail to address effectively.
Innovation Solution
A method and system that schedule charging of an electrified vehicle's energy storage device based on a learned key-on pattern, derived by recursively updating the probability of key-on events, allowing for optimized charging start and end times and energy requirements estimation to match upcoming trips.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If charging starts as soon as external power source is connected, then battery is recharged efficiently, but battery cell capacity and performance deteriorate due to prolonged high state of charge
Solution Approach 1:
The system performs preliminary learning of driver behavior patterns during an acclimation period, building a probabilistic model of key-on events before it is needed for charging control. This allows the system to predict when the driver will next use the vehicle and schedule charging to complete just before that predicted usage, avoiding prolonged high state of charge while ensuring the battery is ready when needed.
Solution Approach 2:
The system continuously monitors actual key-on events and compares them with predicted events, recursively updating the probability model based on the difference between expected and actual behavior. This feedback loop allows the system to adapt to changing driver patterns over time, improving prediction accuracy and optimizing charging schedules to balance battery health with charging efficiency.
2Reliability
If battery is maintained at high state of charge for prolonged periods, then charging readiness is ensured, but battery aging accelerates and overall capacity reduces
Solution Approach 1:
The system transitions from a static charging approach (charge immediately and maintain high state of charge) to a dynamic approach that adapts charging schedules based on learned driver behavior patterns. By making charging schedules dynamic and responsive to actual usage patterns, the system can maintain charging readiness only when necessary, reducing prolonged exposure to high state of charge conditions that cause battery aging.
Solution Approach 2:
The system changes the state of charge parameter dynamically based on predicted usage timing. Instead of maintaining a constantly high state of charge, the system adjusts the charge level and timing to achieve the minimum necessary charge just before predicted key-on events, thereby reducing the duration of high state of charge exposure while ensuring charging readiness when needed.
3Reliability
If charging schedule is optimized based on learned patterns, then battery wear is reduced, but system complexity increases due to learning and prediction algorithms
Solution Approach 1:
The system performs self-learning by automatically monitoring driver behavior and building probabilistic models without requiring external programming or manual input. The system serves itself by using its own operational data to improve its charging control strategy, reducing the need for complex external control systems while achieving adaptive optimization of battery health.
Data Source
AI summary
A method according to an exemplary aspect of the present disclosure includes, among other things, scheduling charging of an energy storage device of an electrified vehicle based on a learned key-on pattern. The learned key-on pattern is derived by recursively updating the probability that a subsequent key-on event is likely to occur at any given time and day.