Three-dimensional magnetic field decoupling and collecting method and system for ion battery

By decoupling the multidimensional time-series data of ion batteries through a neural network model with long short-term memory network and time attention mechanism, the problem of low accuracy of three-dimensional magnetic field signal acquisition in the prior art is solved, and high signal-to-noise ratio magnetic field vector reconstruction is achieved, reducing system complexity and cost.

CN122131149APending Publication Date: 2026-06-02WUHAN TEXTILE UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
WUHAN TEXTILE UNIV
Filing Date
2026-02-06
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

Existing lithium-ion battery monitoring technologies rely on single-dimensional macroscopic parameters, making it difficult to effectively perceive changes in the microscopic current distribution inside the battery. Furthermore, the acquisition of three-dimensional magnetic field signals is affected by environmental noise, temperature drift, and nonlinear coupling relationships, resulting in low accuracy and high cost.

Method used

A neural network model employing long short-term memory and time attention mechanisms is used to achieve high-precision software decoupling of the three-dimensional magnetic field by standardizing and preprocessing multi-dimensional time-series data and jointly optimizing the training, thus avoiding reliance on expensive hardware.

Benefits of technology

Without increasing hardware costs, it significantly improves the accuracy and reliability of three-dimensional magnetic field information acquisition, dynamically adapts to signal characteristics under different working conditions, and achieves high signal-to-noise ratio magnetic field vector reconstruction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122131149A_ABST
    Figure CN122131149A_ABST
Patent Text Reader

Abstract

This disclosure relates to a method and system for decoupling and acquiring a three-dimensional magnetic field in an ion battery. The method includes: acquiring time-series data during the charging and discharging process of the ion battery; performing standardized preprocessing on the time-series data to obtain a preprocessed standardized dataset; constructing an initial neural network model, wherein the initial neural network model includes a long short-term memory network layer, a time attention mechanism layer, and a fully connected output layer connected sequentially; using the preprocessed standardized dataset as training samples, iteratively training the initial neural network model through a joint optimization algorithm to obtain a trained magnetic field decoupling model; acquiring real-time time-series data of the ion battery; and processing the real-time time-series data sequentially through the long short-term memory network layer, time attention mechanism layer, and fully connected output layer in the magnetic field decoupling model to obtain a decoupled three-dimensional magnetic field feature vector. This disclosure can achieve accurate acquisition of a three-dimensional magnetic field.
Need to check novelty before this filing date? Find Prior Art