EV Battery Fast-Charging Control Using Estimated Anode Voltage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for charging electric vehicle batteries lack a direct method to access and estimate anode and cathode voltages, which are crucial for preventing lithium plating and ensuring safe charging operations.
Innovation Solution
A system and method that utilize a reference measurement of the battery during charging to calculate an augmented state, from which anode voltage is determined. This voltage is compared to a threshold, and the charging rate is adjusted accordingly, using techniques such as regression models or neural networks to estimate voltages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct measurement of anode voltage and cathode voltage is performed using a reference electrode, then measurement precision is improved, but device complexity increases due to the need for additional reference electrode access
Solution Approach 1:
The patent creates a virtual copy of the reference electrode's measurement function through machine learning models. Instead of physically accessing the reference electrode, the system trains neural networks or regression models on historical data containing reference electrode measurements, then uses these trained models to predict anode and cathode voltages from readily available terminal voltage and current data, eliminating the need for physical reference electrode access while maintaining measurement accuracy
Solution Approach 2:
The patent introduces machine learning models as an intermediary between available measurements (terminal voltage and current) and the desired measurements (anode and cathode voltages). The models act as a computational mediator that translates easily obtainable electrical parameters into accurate electrode voltage estimates without requiring direct physical access to the reference electrode
2Productivity
If charging rate is increased to improve productivity, then charging speed is improved, but battery degradation increases due to lithium plating
Solution Approach 1:
The patent implements a closed-loop feedback control system where machine learning models continuously predict anode voltage during charging operations. When the predicted anode voltage approaches the lithium plating threshold, the system automatically adjusts the charging rate downward to prevent plating, then can increase the rate again when conditions are safe. This real-time feedback enables aggressive charging when safe while preventing battery damage
Solution Approach 2:
The patent makes the charging rate dynamic rather than static. The system continuously monitors predicted anode voltage and adjusts the charging rate in real-time based on current battery conditions, allowing the charging process to adapt to changing states of charge, temperature, and current levels to maximize speed while maintaining safety
3Measurement precision
If machine learning models are trained on historical data to estimate voltages, then measurement precision is improved, but loss of time occurs during model training and data processing
Solution Approach 1:
The patent performs the computationally intensive model training and data processing in advance, before actual charging operations begin. Historical data from previous charging cycles is collected and used to train the machine learning models offline. Once trained, the models are deployed for real-time voltage estimation during charging, where they make rapid predictions without requiring additional training time. This preliminary action separates the heavy computational work from the time-critical charging operation
Data Source
AI summary
An electric vehicle includes a system for charging a battery of the electric vehicle. The system includes a sensor for obtaining a reference measurement of the battery during charging and a processor. The processor is configured to calculate an augmented state from the reference measurement, determine an anode voltage from the augmented state, compare the anode voltage to a threshold, and adjust a charging rate based on the comparison of the anode voltage to the threshold.