Battery State-of-Charge Correction Using Fleet Aging Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current battery controllers inadequately implement battery aging, leading to inaccurate state-of-charge determinations due to insufficient accuracy in indicating the aging state, which is resource-intensive and challenging to model effectively, especially under dynamic load transitions and recuperation processes.
Innovation Solution
A method that uses a reference state-of-charge model, trained on fleet data, to correct battery model parameters, particularly the series resistor, to adapt the calculated state of charge to a more accurate reference state, improving the precision of state-of-charge estimation by comparing calculated and reference states and updating model parameters based on deviations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a state-of-charge model with battery aging parameters is implemented in the battery controller, then the state of charge can be determined, but the accuracy deteriorates due to insufficient aging state indication
Solution Approach 1:
A central unit is introduced as an intermediary between the battery controller and the aging state model. The central unit receives raw battery data, processes it through a reference aging state model to determine accurate aging state, and sends corrections back to the battery controller. This mediator resolves the contradiction by performing complex aging calculations externally while keeping the battery controller simple.
Solution Approach 2:
The patent replaces the mechanical/computational approach of implementing complex aging models directly in the battery controller with a data-driven approach. The central unit uses machine learning models trained on fleet data to predict aging state, substituting traditional physics-based modeling with statistical learning that achieves higher accuracy without burdening the battery controller.
2Measurement precision
If a reference aging state model is implemented in the battery controller, then aging state can be indicated, but the device complexity increases
Solution Approach 1:
The system is segmented into two independent parts: a simple battery controller that handles real-time state of charge calculations, and a separate central unit that handles complex aging state determination. This segmentation allows each component to be optimized independently - the battery controller remains simple while the central unit performs sophisticated aging analysis.
Solution Approach 2:
The central unit acts as an intermediary that offloads complex aging calculations from the battery controller. Instead of embedding a complex reference aging model in the battery controller, the central unit receives necessary data, performs the complex modeling, and returns corrected parameters, thereby reducing battery controller complexity while maintaining high aging state accuracy.
3Measurement precision
If battery model parameters are corrected using fleet data from a central unit, then state of charge accuracy is improved, but the loss of time occurs during data transmission and correction
Solution Approach 1:
The central unit performs preliminary actions by continuously receiving and processing fleet data to update the reference aging state model in advance. Correction data is prepared and transmitted to vehicles proactively, so when a vehicle needs correction, the data is already available, minimizing the actual correction time and reducing the perceived time loss.
Solution Approach 2:
The system maintains continuous useful action by constantly collecting fleet data, continuously training and updating the reference aging state model, and continuously transmitting corrections to vehicles. This ongoing process ensures that when corrections are applied, they are based on the most current data, maximizing accuracy while the continuous nature of the process distributes the time investment over extended periods.
Data Source
AI summary
A method for operating a system for ascertaining a state of charge of a battery of a motor vehicle includes providing operating variables of the battery, providing a calculated state of charge in the motor vehicle using a state-of-charge model that indicates a calculated state of charge depending on at least one battery model parameter of the state-of-charge model for the battery, and ascertaining a reference state of charge using a reference state-of-charge model depending on the operating variables. The reference state-of-charge model is trained to indicate the reference state of charge depending on the operating variables and/or a predefined aging state. The method further includes performing a correction of the at least one battery model parameter depending on a difference between the reference state of charge and the calculated state of charge in order to adapt the calculated state of charge to the reference state of charge.