Battery Parameter Estimation Using Unscented Kalman Filter

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Mathematical algorithms fail to provide highly accurate estimates of internal battery parameters due to their inadequacy in handling non-linear operational characteristics of batteries.

Innovation Solution

A method and system that determine an estimated battery parameter vector by using a computer to generate predicted battery parameter vectors, state vectors, and output vectors, incorporating uncertainty, to calculate an accurate parameter estimate based on measured values and sensor data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional mathematical algorithms are used to estimate battery parameters, then the device complexity is low, but the measurement precision of battery parameters is insufficient

Engineering Contradiction:

Improvebattery parameter estimation accuracyVSAvoidalgorithm complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent transforms the battery parameter estimation problem by changing the mathematical approach from traditional algorithms to an unscented Kalman filter-based algorithm. This parameter change in the estimation method enables accurate handling of non-linear battery characteristics while providing optimized estimates of state-of-charge, capacity, and internal resistance

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces traditional mathematical estimation mechanisms with a more advanced unscented Kalman filter algorithm. This substitution introduces sigma points and non-linear transformation capabilities, replacing simple mathematical algorithms with a more sophisticated computational approach that better handles battery non-linearity

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If traditional algorithms are used, then the ease of operation is high, but the reliability of battery parameter estimation is low

Engineering Contradiction:

Improvebattery parameter estimation reliabilityVSAvoidalgorithm implementation complexity

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The patent implements feedback mechanisms through the unscented Kalman filter algorithm, which continuously compares predicted battery states with actual measurements and adjusts estimates accordingly. This feedback loop significantly improves the reliability of battery parameter estimation by accounting for non-linear characteristics and measurement uncertainties

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent performs preliminary actions by pre-defining sigma points and their associated weights before the actual estimation process. This preliminary setup of the unscented Kalman filter algorithm enables more accurate prediction of battery states by anticipating non-linear behavior patterns

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentEP1971872B1System, method, and article of manufacture for determining an estimated battery parameter vector

Publication Date: 2018.09.26 LG CHEM LTD

Figure 1
Figure 2
Figure 3

AI summary

A system, a method, and an article of manufacture for determining an estimated battery parameter vector indicative of a parameter of a battery are provided. The method determines a first estimated battery parameter vector indicative of a parameter of the battery at a first predetermined time based on a plurality of predicted battery parameter vectors, a plurality of predicted battery output vectors, and a first battery output vector.