Research And Application Of Coal Flow Identification Technology For Belt Conveyor

With the development of China’s coal industry,underground coal transportation tends to increase speed and belt.The belt conveyor has become the main tool for coal transportation under the mine.However,in the process of coal transportation,especially at the coal drop between the belts,when the coal falls,it is prone to abnormal situations such as block of gangue,plugging and tearing of the belt with anchor rods,etc.,causing damage to the equipment and downhole Hidden dangers safely.The traditional methods for identifying coal flow anomalies in belt conveyors are mainly manual detection and sensor detection,but manual detection is time-consuming and labor-intensive,and is prone to negligence;sensor detection is mostly based on contact methods,which are affected by the humid and complex environment under the mine.In view of the above problems,this paper conducts the research on coal conveyor flow recognition technology from the perspective of image processing,proposes gangue detection and bolt detection algorithms,designs the conveyor conveyor coal flow detection and recognition system,and tests this method at an industrial site.The main innovations of this article are as follows:(1)Research on the method of using video image analysis for downhole target recognition,which provides a new idea for the detection algorithm of gangue of belt conveyor.First,the color video image is converted into a grayscale image by preprocessing,and Gaussian smoothing is performed to reduce the algorithm complexity and environmental noise.Then,in view of the "ghosting" problem and poor detection caused by background disturbances in complex environments existing in the traditional Vi Be algorithm in coal flow recognition,a gangue detection algorithm based on Vi Be is proposed from two aspects of improved background modeling initialization method and adaptive threshold.Finally,calculate the relative area of the moving target and compare it with the alert value to determine if there is a large piece of vermiculite in the picture.Experiments show that the proposed gangue detection algorithm is simple to calculate and easy to implement,and can effectively detect the large gangue appearing in the video picture.(2)Aiming at the anchors that may exist in the video picture,according to their characteristics,an anchor detection algorithm suitable for underground conveyors is proposed.On the basis of preprocessing,the current motion state is analyzed using the inter-frame difference method to determine whether the current frame is a coal transport image,eliminate the influence of belt idling and still pictures on the detection results.,the monitoring image to meet the requirements,with Canny edge information extraction algorithm,and do cumulative probability Hough transform,the recording ends of the lines in the picture and computing the length of the anchor with a preset determination threshold value,it is determined whether there is a anchor screen.Experimental results show that the proposed anchor detection algorithm can effectively identify the anchor in the video monitor screen,and meet real-time requirements.(3)Use C ++ language combined with Open CV open source library to design a belt conveyor coal flow recognition system,and carry out industrial tests on the method proposed in this paper.The system can load and process 4 video monitoring terminals in real time,intelligently detect the gangue and anchors that may exist in the video screen,and also have auxiliary functions such as coal pile detection and coal quantity monitoring.When an abnormality is found,it can send an abnormal signal to the network relay to realize linkage control.The system also supports custom detection thresholds to meet the needs of different scenarios.There are 39 figures,9 tables,and 80 references in this paper.