Battery Degradation Estimation from Partial Charge-Discharge Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating storage battery degradation, such as coulomb counting, SOC-OCV curve method, and machine learning, face challenges like requiring full charge cycles, temperature variations affecting SOC-OCV curves, and difficulty in real-time estimation while the battery is in use.
Innovation Solution
A storage battery degradation estimation device and method that calculates voltage and electric charge change amounts during charging and discharging, using these values as inputs for a machine learning estimation model to estimate degradation state, with optional low-pass filtering for smoothing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coulomb counting method is used to calculate capacity retention rate, then the full charge capacity can be calculated by integrating charge current, but the storage battery must be fully charged from completely discharged state which is inconvenient and difficult when device is always in operation
Solution Approach 1:
The patent pre-calculates and stores SOC-OCV correspondence data during battery manufacturing or initial setup, creating a lookup table that maps state of charge values to open circuit voltages. This preliminary action eliminates the need for full charge cycles during operation, as the device can directly reference pre-established data to determine capacity retention rate from instantaneous OCV measurements.
2Measurement precision
If SOC-OCV curve method is used to calculate capacity retention rate, then full charge capacity can be calculated from open circuit voltage measurement, but calculation errors occur due to temperature variations affecting SOC-OCV curve properties
Solution Approach 1:
The patent stores multiple SOC-OCV correspondence datasets corresponding to different temperature conditions (e.g., -30°C, -10°C, 10°C, 30°C, 50°C). When temperature varies, the system selects or interpolates between appropriate datasets to compensate for temperature-induced curve shifts, thereby maintaining calculation accuracy across varying thermal conditions without requiring real-time curve recalibration.
3Measurement precision
If machine learning method is used to estimate capacity retention rate, then degradation can be estimated using various battery states as inputs, but measured data of multiple full charges are required which makes immediate calculation difficult while battery is in use
Solution Approach 1:
The patent enables capacity retention rate estimation using partial charge-discharge cycle data rather than requiring complete full charge cycles. The machine learning model is designed to process and extract degradation information from incremental measurements taken during normal operational cycles, allowing immediate estimation without waiting for excessive data accumulation from multiple full charges.
4Measurement precision
If multiple SOC-OCV curves are switched to address property variations, then temperature effects can be compensated, but calculation errors occur due to erroneous switching between curves
Solution Approach 1:
The patent introduces temperature as an intermediary parameter that directly indexes into the appropriate SOC-OCV correspondence dataset. Instead of complex switching logic that compares multiple curves and determines optimal selection, the system uses temperature measurement to directly select or interpolate between pre-stored datasets, simplifying the selection process and eliminating erroneous switching while maintaining temperature compensation benefits.
Data Source
AI summary
A degradation estimation device includes: a first calculator that calculates a voltage change amount from a first voltage value and a second voltage value measured when the storage battery is charged and discharged; a second calculator that calculates an electric charge change amount from a first current value and a second current value when the storage battery is charged and discharged; a data storage that stores measured data that includes the voltage change amount and the electric charge change amount; a model storage that stores an estimation model that uses one or more electric charge change amounts as inputs and outputs the degradation state of the storage battery; and an estimator that estimates the degradation state of the storage battery by using the estimation model and using, as inputs, the one or more voltage change amounts and the one or more electric charge change amounts.