Battery State Estimation Data Filtering

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing battery management systems face challenges in accurately estimating the state of charge (SOC) and state of health (SOH) of batteries due to variations in charging voltage data over time, leading to potential errors in abnormality detection and resource inefficiency.

Innovation Solution

A battery state estimation method and apparatus that preprocesses sensing data to select suitable data for estimation based on feature information such as data size and variance, using a controller to filter and downsample data, and determine state information either from selected or previous data, thereby optimizing resource usage and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If all sensing data is used for battery state estimation, then more data is available for calculation, but computational resources are wasted and accuracy decreases due to unsuitable data

Engineering Contradiction:
Improvebattery state estimation accuracyVSAvoidcomputing resource consumption
Core Design Contradiction:
Measurement precisionVSLoss of energy

Solution Approach 1:

The patent divides sensing data into multiple segments based on time intervals and selects only suitable segments for state estimation. This segmentation approach filters out unsuitable data (such as data during abnormal operations or insufficient charging/discharging cycles) while retaining valuable data, thereby improving estimation accuracy and reducing unnecessary computational resource consumption.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter of data selection by evaluating multiple criteria including data length, time interval, and operational conditions before determining whether sensing data is suitable for estimation. This parameter-based filtering ensures that only data meeting specific thresholds is processed, optimizing both accuracy and resource efficiency.

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If sensing data with insufficient length is used for estimation, then more data points are available for calculation, but estimation accuracy decreases

Engineering Contradiction:
Improvedata quantityVSAvoidstate estimation accuracy
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent performs preliminary evaluation of sensing data before using it for state estimation. It checks whether the data length meets minimum requirements and whether the data was collected under appropriate conditions (normal operation, sufficient charging/discharging cycles). This preliminary action prevents inaccurate estimation by filtering out insufficient data before the estimation process begins.

Inventive Principle:
Principle #10Preliminary action

3Quantity of substance

If data collected during abnormal battery operation is used for estimation, then more data is available, but estimation accuracy deteriorates

Engineering Contradiction:
Improvedata quantityVSAvoidestimation reliability
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent introduces additional parameters for data evaluation, including operational status flags that indicate whether the battery was operating normally during data collection. By changing the selection criteria to include these reliability parameters, the system filters out data collected during abnormal conditions (such as overheating, excessive current, or error states) while maintaining sufficient data quantity from normal operations.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3349026B1Method and apparatus for estimating state of battery
Publication Date: 2023.05.10 SAMSUNG ELECTRONICS CO LTD
  • EP3349026B1 patent drawingFigure 1~2
  • EP3349026B1 patent drawingFigure 3~4
  • EP3349026B1 patent drawingFigure 5~6

AI summary

A battery state estimation method and apparatus (100) are provided. Sensing data of a battery is received, and feature information is acquired by preprocessing the sensing data. The preprocessed sensing data is selected based on the feature information, and state information of the battery is determined based on at least one of the selected preprocessed sensing data or previous state information of the battery.