Lithium ion battery capacity prediction method based on statistical characteristic evolutionary coding

By constructing a statistical feature-based evolutionary coding method in lithium-ion battery health state modeling and using a genetic algorithm to generate combinatorial mathematical expressions, the problems of insufficient nonlinear expression capability and high computational resources in existing technologies are solved, and efficient and interpretable lithium-ion battery capacity prediction is achieved.

CN120742115BActive Publication Date: 2026-07-21CHONGQING UNIV
2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHONGQING UNIV
Filing Date
2025-07-21
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

Existing lithium-ion battery health status modeling methods suffer from several problems when dealing with long-term complex degradation modes. These problems include difficulty in expressing nonlinear evolution laws, underutilization of high-order combination relationships of original features, limited model generalization ability, and high computational resource requirements, making it difficult to meet the requirements of real-time performance and reliability.

Method used

A statistical feature-based evolutionary coding method is adopted. A genetic algorithm is used to construct a combinatorial mathematical expression in the original feature space to form coded features. A linear regression model is combined for capacity prediction. The feature combination is represented by a full binary tree or a partial binary tree structure. The high-order interaction relationship between features is automatically mined to construct a lightweight nonlinear prediction model.

Benefits of technology

It improves the accuracy and stability of lithium-ion battery capacity prediction, has good computational efficiency and deployment flexibility, is suitable for practical application environments with limited data scale or limited computing resources, and has interpretability and visualization analysis capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120742115B_ABST
    Figure CN120742115B_ABST
Patent Text Reader

Abstract

The present application relates to a kind of lithium ion battery capacity prediction method based on statistical characteristic evolution coding, belong to lithium ion battery health state estimation technical field.The method extracts voltage and current statistical characteristics in the process of battery charging, uses correlation analysis to filter original features, and constructs coding feature expression based on genetic algorithm with nonlinear expression ability, combined with original features to input regression model to predict capacity.Compared with traditional methods, the present application improves the nonlinear expression ability of the model while retaining the interpretability of the features, has higher prediction accuracy and stronger generalization ability, and is suitable for efficient estimation of battery capacity in engineering environments with limited computing resources.
Need to check novelty before this filing date? Find Prior Art