A supercapacitor series module early fault identification method based on monomer difference deduction model

By constructing a single-unit difference inference model and using Z-score standardization and local outlier factor methods, the problems of computational resource limitations and misjudgment in early fault identification of supercapacitor modules are solved, achieving efficient and accurate fault identification and early warning.

CN122283518APending Publication Date: 2026-06-26TONGJI UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
TONGJI UNIV
Filing Date
2026-03-25
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

In supercapacitor modules, the electrical characteristics of early faults change slightly and are easily affected by noise. Existing methods are difficult to effectively identify in vehicle systems with limited computing resources, and traditional methods are prone to misjudgment, posing a safety risk.

Method used

A single-unit difference inference model is constructed. By identifying parameters and estimating the state of charge of a small number of reference units, early faults are identified using Z-score normalization and local outlier factor methods, reducing computational consumption and amplifying fault characteristics.

Benefits of technology

It achieves efficient early fault identification of supercapacitor modules under limited computing resources, improves the sensitivity and accuracy of the identification algorithm, and can provide early warning of potential faults.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122283518A_ABST
    Figure CN122283518A_ABST
Patent Text Reader

Abstract

This invention proposes an early fault identification method for supercapacitor series modules based on a single-unit difference inference model. Addressing the issue of fault identification relying on massive computational resources, this invention constructs a single-unit difference inference model by deriving the state and parameter differences between individual units. This model allows for the estimation of the operating state of the remaining non-reference units using only a small amount of information from reference units, achieving low-computational-power estimation of the entire module's state. Furthermore, Z-score normalization amplifies subtle early fault anomalies, and a local outlier factor algorithm is used to detect and locate faulty units, enabling early connection and short-circuit fault identification. This method eliminates the need for modeling and identifying all units, significantly reducing computational complexity and improving fault identification efficiency. Simultaneously, it utilizes state parameters from the supercapacitor module's SOC estimation process to extract early, weak fault features, eliminating the need for fault model building and further enhancing system computational efficiency. This method possesses significant engineering application value and promising prospects for wider application.
Need to check novelty before this filing date? Find Prior Art