摘要
"绿色、安全、和谐、智能、高效"成为矿业可持续发展的时代要求,智能矿山的建设有助于提高产业的自动化和智能化水平。井下电机车无人驾驶是井下运输智能化无人技术应用的重要一环,轨道障碍物检测作为无人驾驶的关键技术,能够保障井下有轨运输的效率和安全。为了达到不同距离、不同位置轨道中间和两侧的人、设备、碎石的识别和预警的目的,分析电机车的运行条件,梳理轨道障碍物的类型,根据运输环境和计算机视觉技术,对电机车在井下运输过程中影响运行的轨道障碍物进行智能识别。结合传统图像处理方法的轨道定位技术,描述电机车轨道区域划分的方法。采用最新的深度学习目标检测算法YOLOv5,对影响电机车运行的诸多因素分析和判断,是图像处理在矿山智能化的创新应用。
"Green,Safe,Harmonious,Intelligent,Efficient"has become the era requirement of sustainable develop⁃ment of mining industry.The construction of intelligent mine helps to improve the automation and intelligent level of the in⁃dustry.Underground electric locomotive driverless is an important part of underground transportation intelligent unmanned technology application.As the key technology of driverless,track obstacle detection can ensure the efficiency and safety of underground rail transportation.In order to achieve the purpose of identification and early warning of people,equipment and gravel in the middle and both sides of the track at different distances and positions,the operation conditions of the electric lo⁃comotive are analyzed,the types of track obstacles are sorted out,and the track obstacles that affect the operation of the elec⁃tric locomotive in the underground transportation process are intelligently identified according to the transportation environ⁃ment and computer vision technology.Combined with the track positioning technology of traditional image processing meth⁃od,the method of track area division of electric locomotive is described.Using the latest deep learning target detection algo⁃rithm YOLOv5 to analyze and judge many factors affecting the operation of electric locomotive is an innovative application of image processing in mine intelligence.
作者
于骞翔
张元生
YU Qianxiang;ZHANG Yuansheng(BGRIMM Technology Group,Beijing 102628,China;Beijing Key Laboratory of Nonferrous Intelligent Mining Technology,Beijing 102628,China;BGRIMM Intelligent Technology Co.,Ltd.,Beijing 102628,China)
出处
《金属矿山》
CAS
北大核心
2021年第8期150-157,共8页
Metal Mine
基金
“十三五”国家重点研发计划项目(编号:2018YFC0604404)。