Study On Gas Leak Identification Method Of Compressor Pipeline In Coal Mine

Mine air compressor is an important guarantee equipment for coal mine safety production.Compressed air is transported to the underground through pipelines.On the one hand,compressed air is used to provide power for pneumatic tools such as down-hole hammers and air pumps,on the other hand,it could ensure that there is enough oxygen underground and guarantee the safety of workers’ lives.Due to the high-pressure characteristics of compressed gas,leaking failures often occur during transportation of compressed gas.At present,for the detection of such faults,it is mainly to judge whether the pipeline has leaked through the listening and watching of the patrolling personnel.This manual patrol method has low detection efficiency and poor real-time performance,resulting in wasting of energy and safety accidents in coal mine production from time to time.In addition,the temperature and noise of the compressor room are high,and the working environment of the workers is harsh.In order to achieve real-time monitoring of gas leak faults in mine air compressor pipelines,this paper obtain infrared images of leaking pipelines based on infrared thermal imaging technology,use improved wavelet threshold algorithm optimized by firefly algorithm for image denoising,and use Otsu-Grabcut algorithm to segment the image,combine the directional gradient histogram and the gray level co-occurrence matrix to extract the image features,and use the improved firefly algorithm to optimize the support vector machine to realize the air compressor gas delivery pipeline leak fault recognition.The main research contents of the paper are as follows:(1)Analyzed the basic structure of mine air compressors,studied the main causes of gas pipeline leak,proposed the functional requirements of the gas leak identification system,and designed the architecture and identification process of the air compressor pipeline gas leak identification system based on infrared thermal imaging technology.(2)Analyzed the temperature characteristics of the gas delivery pipeline of the mine air compressor,constructed a three-dimensional model of the gas pipeline leak.Proposed an infrared image denoising algorithm for pipeline gas leak based on a combination of improved firefly algorithm and wavelet threshold algorithm,studied the Otsu-Grabcut infrared image segmentation algorithm,and verified the effectiveness of the above models and algorithms through simulation analysis.(3)Designed a method for extracting infrared image features of gas leak of air compressor pipelines on mine based on directional gradient histogram and gray level co-occurrence matrix,and optimized support vector machine penalty factors and kernel function parameters based on improved firefly algorithm.Used the confusion matrix as the evaluation index of the support vector machine classification result,to evaluate the classification accuracy of the support vector machine.(4)Built an experimental platform for pipeline gas leak identification for mine air compressors,and used the platform to test the leak and wear of the pipeline and the leak failure of the valve.Tested the function of the gas leak identification system for mine air compressor pipeline,and verified the effectiveness of the gas leak identification method for mine air compressor pipeline studied in this paper.The recognition accuracy is about 96%,meeting the needs of the site.There are 33 figures,19 tables and 94 references in this paper.