Scholarlink

<p>With the continuous improvement of the automation level of coal mining, more and more electrical equipment are entering coal mines. Traditional manual inspection and fault diagnosis methods are difficult to meet the safety needs of coal mining. Therefore, this paper designs an intelligent fault diagnosis system for electrical equipment in coal mines based on computer vision and deep learning (DL). This study proposes a novel integrated framework that synergizes the global search capability of GWO (Gray Wolf Optimization)-OTSU for noise-robust segmentation, the precision of KNN (K-Nearest Neighbors) for non-linear temperature mapping, and the efficiency of LF-YOLOv7 (Lightweight and Fast-You Only Look Once version 7) for real-time detection, thereby overcoming the limitations of isolated diagnostic methods in complex underground environments.This system collects image data of target electrical equipment through infrared thermal imagers or infrared cameras, and uses the optimized OTSU algorithm based on GWO to segment the image. It uses the KNN algorithm to extract features of temperature values, and finally uses the DL model based on LF-YOLOv7 to identify faults and warn of temperature features. Compared with the traditional EfficientDet, SSD, YOLOX-s and YOLOv7 models, the system in this paper has better performance, reaching 18.97&#xa0;M parameters, 90.63FPS, and 98.97% mean average precision (mAP). The YOLOv7 model reached 46.98&#xa0;M parameters and 72.11FPS, and the mAP was only 89.31%. This system provides a convenient and reliable method for underground electrical equipment detection, and ensures safety.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Design of intelligent fault diagnosis system for electrical equipment in coal mines based on computer vision and deep learning

  • 期刊论文

  • 发表日期：

    2026年3月13日

查看全文
  • Qinzhu Wang

摘要

With the continuous improvement of the automation level of coal mining, more and more electrical equipment are entering coal mines. Traditional manual inspection and fault diagnosis methods are difficult to meet the safety needs of coal mining. Therefore, this paper designs an intelligent fault diagnosis system for electrical equipment in coal mines based on computer vision and deep learning (DL). This study proposes a novel integrated framework that synergizes the global search capability of GWO (Gray Wolf Optimization)-OTSU for noise-robust segmentation, the precision of KNN (K-Nearest Neighbors) for non-linear temperature mapping, and the efficiency of LF-YOLOv7 (Lightweight and Fast-You Only Look Once version 7) for real-time detection, thereby overcoming the limitations of isolated diagnostic methods in complex underground environments.This system collects image data of target electrical equipment through infrared thermal imagers or infrared cameras, and uses the optimized OTSU algorithm based on GWO to segment the image. It uses the KNN algorithm to extract features of temperature values, and finally uses the DL model based on LF-YOLOv7 to identify faults and warn of temperature features. Compared with the traditional EfficientDet, SSD, YOLOX-s and YOLOv7 models, the system in this paper has better performance, reaching 18.97 M parameters, 90.63FPS, and 98.97% mean average precision (mAP). The YOLOv7 model reached 46.98 M parameters and 72.11FPS, and the mAP was only 89.31%. This system provides a convenient and reliable method for underground electrical equipment detection, and ensures safety.