基于YOLO-CDBW模型的列车接触网异物检测研究  

Research on Foreign Object Detection in Railway Overhead Contact System Based on YOLO-CDBW Model

作　　者：郭翔羽 石天怡 陈燕楠 南新元[1] 蔡鑫[1] GUO Xiangyu;SHI Tianyi;CHEN Yannan;NAN Xinyuan;CAI Xin(School of Electrical Engineering,Xinjiang University,Urumqi Xinjiang 830017,China)

机构地区：[1]新疆大学电气工程学院,新疆乌鲁木齐830017

出　　处：《广西师范大学学报(自然科学版)》2025年第2期56-69,共14页Journal of Guangxi Normal University:Natural Science Edition

基　　金：国家自然科学基金(62303394);新疆维吾尔自治区自然科学基金(2022D01C693);新疆维吾尔自治区高校基本科研业务费科研项目(XJEDU2023P025)。

摘　　要：接触网是为列车行驶提供电力的输电线路,附着在接触网的塑料袋等异物会对列车运行造成安全隐患。针对目前人工检查效率低下、劳动成本高等问题,本文提出一种基于YOLOv7改进的接触网异物检测模型YOLO-CDBW。首先,在特征提取阶段构建一种使用残差瓶颈结构和深度分离卷积层的特征提取模块,避免因网络深度增加造成的小目标特征丢失问题,并降低网络运算量;其次,颈部改用BiFPN结构,捕捉多尺度信息,改善细节特征丢失问题,同时嵌入BiFormer注意力机制,重新分配融合后特征图的权重,提高网络对异物的关注度;最后,使用WIoU损失函数优化模型,通过动态聚焦机制,将注意力聚集在普通质量锚框上,提高预测精准度。经实验,YOLO-CDBW模型平均精度均值mAP 0.5达到87.1%,检测速度FPS达到66.5 frame/s,较YOLOv7模型分别提高5.0和10.8个百分点,满足接触网异物检测需求。The catenary is a transmission line that provides power for the train,and foreign objects such as plastic bags attached to the catenary will cause potential safety hazards to the train operation.In order to solve the problems of low efficiency and high labor cost of manual inspection,a YOLO-CDBW model for catenary foreign body detection based on YOLOv7 is proposed.Firstly,in the feature extraction stage,a feature extraction module using residual bottleneck structure and depth separation convolutional layer is constructed to avoid the problem of small target feature loss caused by the increase of network depth and reduce the amount of network computation.Finally,the WIoU loss function is used to optimize the model and focus on the ordinary mass anchor frame through the dynamic focusing mechanism to improve the prediction accuracy.Experimental results show that,the average mAP 0.5 of the YOLO-CDBW model reaches 87.1%and the detection speed FPS reaches 66.5 frame/s,which are 5.0 and 10.8 percentage points higher than those of the YOLOv7 model,respectively,meeting the needs of catenary foreign body detection.

关 键 词：接触网 异物检测 YOLO 目标检测 损失函数 注意力机制 

分 类 号：U226.8[交通运输工程—道路与铁道工程] TP391.41[自动化与计算机技术—计算机应用技术]