图像处理技术在无人综采系统中的应用研究  

Application of image processing technology in unmanned fully mechanized coal mining system

作　　者：闫旭 邬建雄 YAN Xu;WU Jianxiong(China Energy Group Shendong Coal Group Co.,Ltd.,Ordos 017000,China)

机构地区：[1]国能神东煤炭集团有限责任公司,内蒙古鄂尔多斯017000

出　　处：《煤炭科学技术》2022年第S01期269-276,共8页Coal Science and Technology

摘　　要：煤炭资源在我国能源体系中是不可替代的,煤炭工业的持续、安全、健康、绿色的开采是煤炭行业的发展方向,但煤炭开采属高危行业,因此开展无人开采工作面研究具有重要意义。受到煤层结构、煤尘、震动、噪声以及相关技术未突破等因素影响,煤层状态难以预判,采煤机滚筒在截割过程中经常遇到夹矸和岩石断层,工作面无人化割煤难以真正实现。而采煤工人却可以通过视觉、听觉的判断,控制采煤机作业。因此从视觉判断的技术角度,研究利用图像处理技术实现人眼判断工况的方法具有可行性。利用图像的特征向量、相似度判断、检索平台应用等技术,可以实现图像的检索,采煤机工作过程中,其姿态和位置可以理解成是由多张图片组成,建立其工作的图像库,并在工作过程中不断更新、丰富图库,当积累到一定程度,将图库分成正常姿态、割顶、割底、报警等各类图像,全覆盖采煤机的各类姿态,通过专家系统进行检索当前状态与图库中存储的状态进行对比,通过相似的数值来确定姿态,进而达到采煤司机视觉判断的效果,解决目前传感器无法准确预测采煤机工作姿态、煤岩识别等技术难题,实现采煤机的自学习、自诊断,自控制开采。Coal resources are irreplaceable in China′s energy system.The sustainable,safe,healthy and green mining of coal industry is the development direction of coal industry,but coal mining is a high-risk industry.Therefore,it is of great significance to carry out the research on unmanned mining face.Affected by the coal seam structure,coal dust,vibration,noise and the failure of relevant technologies,it is difficult to predict the coal seam state.The shearer drum often encounters gangue and rock faults in the cutting process,and it is difficult to realize unmanned coal cutting in the working face.However,coal miners can control the operation of shearers through visual and auditory judgment.Therefore,from the perspective of visual judgment technology,it is feasible to study the method of using image processing technology to judge the working condition by human eyes.Image retrieval can be realized by using image feature vector,similarity judgment,retrieval platform application and other technologies.In the working process of shearer,its posture and position can be understood as composed of multiple pictures.Establish its working image database,and constantly update and enrich the image database in the working process.When it is accumulated to a certain extent,the image database is divided into normal posture,top cutting,bottom cutting,alarm and other images,All kinds of attitudes of the full coverage shearer are retrieved through the expert system,and the current state is compared with the state stored in the drawing library.The attitude is determined through similar values,so as to achieve the effect of visual judgment of the shearer driver,solve the technical problems that the current sensors can not accurately predict the working attitude of the shearer and coal rock identification,and realize the self-learning,self diagnosis and self-controlled mining of the shearer.

关 键 词：图像处理 图像检索 煤岩识别 无人工作面 采煤机 

分 类 号：TD632.1[矿业工程—矿山机电]