基于RT-DETR算法的皮带运输机异物识别方法  

Recognition Method of Foreign Objects in Belt Conveyor Based on RT-DETR Algorithm

作　　者：杨家威 Yang Jiawei(Shanxi Coal International Energy Group Corporation,Taiyuan Shanxi 030000,China)

机构地区：[1]山煤国际能源集团股份有限公司,山西太原030000

出　　处：《机械管理开发》2025年第2期183-185,共3页Mechanical Management and Development

摘　　要：为提高皮带运输机在工业生产中的安全性,基于RT-DETR算法的异物识别方法,进行了RT-DETR模型架构,并设计了皮带运输机异物识别系统,通过实时数据采集模块、数据传输与处理模块、实时推理模块、异物检测与报警模块、用户界面模块等功能模块可实现皮带运输机异物高效识别,通过对比实验测试,结果表明,该系统能提高识别速度与准确率,有效减少因异物引起的设备故障与生产事故,具有广泛的应用前景。In order to improve the safety of belt conveyor in industrial production,this study adopts the foreign object recognition method based on RT-DETR(Real-Time Detection Transformer)algorithm,carries out the RT-DETR model architecture,and designs the foreign object recognition system of the belt conveyor,which passes through the real-time data acquisition module,data transmission and processing module,Through real-time data acquisition module,data transmission and processing module,real-time inference module,foreign object detection and alarm module,user interface module and other functional modules can realize the efficient identification of foreign objects in belt conveyor,through the comparative experimental test,the results show that this system can improve the identification speed and accuracy,effectively reduce the equipment faults and production accidents caused by foreign objects,and has a wide range of application prospects.

关 键 词：RT-DETR算法 皮带运输机 异物识别系统 

分 类 号：TD528.1[矿业工程—矿山机电] TP391.41[自动化与计算机技术—计算机应用技术]