一种基于声纹检测的储能系统电弧故障诊断方法及系统
By using voiceprint detection technology, combined with multi-branch convolutional neural networks and temporal convolutional networks, and utilizing multi-head temporal attention networks, sensitive identification and accurate hierarchical localization of arc faults in energy storage systems are achieved. This solves the problems of high false alarm rate and false negative rate in existing arc fault detection technologies and improves the safety early warning capability of energy storage systems.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 深圳晶锶科创有限公司
- Filing Date
- 2025-04-29
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for arc fault detection in energy storage systems suffer from insufficient sensitivity, susceptibility to load fluctuations, and difficulty in resolving the dynamic characteristics of arc time-frequency domains through fixed-band spectrum analysis. Shallow machine learning models lack anti-interference capabilities and time-series feature modeling capabilities, resulting in high false alarm rates, missed detection rates, and a lack of fault classification and location capabilities, thus failing to meet the requirements for high-precision safety early warning.
A voiceprint detection-based method is adopted, which collects voiceprint signals through an acoustic sensor array, constructs a multi-branch convolutional neural network and a temporal convolutional network, and combines them with a multi-head temporal attention network to identify the high-frequency spike pulse and low-frequency discharge howling characteristics of the electric arc, perform fault type identification, severity classification and sound source localization, and generate alarm signals and control commands.
It enables sensitive identification, precise classification, and accurate location of arc faults, improves detection sensitivity and anti-interference capabilities, supports fault severity classification and precise sound source location, and enhances the safety protection level and fault early warning capability of energy storage systems.
Smart Images
Figure CN120472932B_ABST