Battery Model Correction via Twin Model Predictive Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current electrochemical energy storage systems face issues such as inconsistent cell balancing, inaccurate State of Charge (SOC) estimation, severe battery health attenuation, frequent safety accidents, and short lifespan due to variations in battery cell characteristics and integration levels among different manufacturers.
Innovation Solution
A correction method and device for an energy storage battery management system that employs a twin model deployed on a server, generating predictive data from historical battery data to train generic battery models and issue model updates to local managers, ensuring consistent battery model updates and improved management of energy storage battery clusters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If generic battery models are used for energy storage battery management, then device complexity is reduced, but manufacturing precision and reliability deteriorate due to inconsistent cell balancing and inaccurate SOC estimation
Solution Approach 1:
The patent changes the parameters of the battery model by training generic battery models with historical battery data to obtain corrected models that adapt to specific battery characteristics. This allows the system to maintain simplicity while improving precision through parameter optimization rather than structural complexity
Solution Approach 2:
The patent creates corrected battery models that copy and adapt generic models to specific battery instances by training them with historical data. This copying process preserves the simplicity of generic models while incorporating specific characteristics for improved accuracy
2Measurement precision
If battery models are frequently updated to improve accuracy, then measurement precision improves, but loss of time increases due to continuous model correction and firmware updates
Solution Approach 1:
The patent performs preliminary actions by collecting and storing historical battery data in advance, which is then used to train corrected models. This preliminary data collection enables faster model correction when needed, reducing the time loss during actual updates
Solution Approach 2:
The patent implements feedback mechanisms where the battery management system continuously monitors battery performance and compares it with model predictions. This feedback enables targeted model corrections only when deviations are detected, reducing unnecessary update time while maintaining precision
3Adaptability or versatility
If multiple generic battery models are deployed on server for comprehensive coverage, then adaptability improves, but device complexity and energy consumption increase
Solution Approach 1:
The patent makes the battery management system universal by training generic battery models with diverse historical data from different battery types and manufacturers. This enables a single corrected model structure to adapt to multiple battery types without requiring separate specialized models for each type
Solution Approach 2:
The patent achieves adaptability through parameter changes rather than structural changes. By training generic models with historical data from various battery types, the system adjusts model parameters to fit different battery characteristics, maintaining a unified model framework while covering diverse battery types
Data Source
AI summary
Embodiments of the present disclosure provide a correction method for an energy storage battery management system, comprising: generating (S210) predictive data based on historical battery data by means of a twin model; training generic battery models according to the predictive data when a model correction event is detected, to obtain (S220) a target battery model; and issuing (S230) model update firmware or model update parameters of the target battery model to a local energy storage battery manager. The present disclosure further provides a correction device for an energy storage battery management system, and a system and a medium.