Research On Intelligent Coal Conveyor Detection In Underground Coal Mine

Coal is an important energy industry in China,which supports economic and social development.At present,due to the increasing demand for coal and the traditional production mode,it often causes the secondary waste of energy,low coal mining rate and high accident rate.Especially in the process of transporting coal with conveyor belt,the waste of electric energy caused by idling of conveyor belt,uneven distribution of coal on conveyor belt and long-term overload operation have caused damage to conveyor,which are still the main problems faced by coal mining enterprises.A problem that coal enterprises need to solve is how to control the speed and start and stop about the conveyor belt by the amount of coal on the conveyor belt at present.Relying on the electronic monitoring system of coal mine,this paper studies the problems above.The major innovation points,algorithm principle and experimental results are as follows:1.A no-load detection method based on conveyer belt in underground coal mine is proposed.Principle of the algorithm: In order to save the power loss caused by the downhole conveyer belt no-load,this paper proposes a conveyer belt no-load detection method combining edge structure similarity algorithm and YOLOV3.The edge structure similarity algorithm was used to fuse the structural features and edge features,and the similarity of each adjacent 10 frames was compared to judge the running state of the conveyor belt for three consecutive times.If the conveyor belt is in operation,the YOLOV3 model with adaptive anchor frame mechanism is used to detect the amount of coal on the conveyor belt and finally determine whether the conveyor belt is empty.The experimental results show that this method can effectively and accurately judge the no-load state of the conveyor belt,and the detection accuracy rate reaches95.8%.2.A method for detecting coal content of conveyor belt based on RES2-UNET is proposed.The principle of the algorithm: through the RES2-UNET model to obtain salient information,the fusion of gray scale,texture,edge and other features into a single network,to achieve the conveyor belt coal volume detection.Finally,the SVM model was established to classify the coal quantity by using the pixel area,length-width ratio and gray value of the coal quantity image of the conveyor belt as the input of the model.And we can see that the classification accuracy of all kinds of coal is more than 93%from the experiment,which can more accurately classify coal quantity,verifying the effectiveness of the model for coal quantity classification.3.An intelligent coal conveyer system for underground coal mine has been developed.The intelligent coal conveying system designed and completed in this paper is realized in MATLAB software.By building the binding of software interface architecture diagram and function file,the corresponding detection algorithm is called,so as to realize the coal quantity detection of conveyer belt in underground coal mine.This system greatly facilitates the operation and management of coal quantity detection.