Research On Leak Detection And Location Algorithm Of Branch Pipeline Based On Fluid Acoustic Emission

In the production of coal mines,the underground pressure wind and water supply pipeline system provides the functions of dust removal,ventilation and oxygen supply for underground operation,which is indispensable for safety production.However,due to the numerous branches,wide distribution area and leakage of pipes in coal mine,how to conduct integral real-time monitoring to the complex pressure wind and water supply system has become an urgent problem in the production of coal mines.According to the characteristics of underground branch pipeline,a leak detection and location algorithm based on acoustic emission was presented,and the main research work was as followsFirstly,combined with the multi-branch structure of underground pipeline,the methods of fluid acoustic emission and changepoint detection were introduced into the leakage detection of branch network.The acoustic emission sensors distributed in different positions were used to monitor the vibration signal of the tube wall in real time,and the statistical characteristics of the variance of the signals were detected by the algorithm of changepoint detection algorithm,and the abnormal tube segmentscould be determined.The interference of underground pipeline's impact and waterhammer on acoustic emission signal was eliminated by changepoint detection adding time threshold method.Secondly,leak detection and location were conducted by using the down-sampling data and first-time sampling data respectively,so as to ensure the high real-time performance of the system.The acoustic emission sensor could collect the vibration signal of tube wall in real time,and the underground substation could sample and store the acoustic emission signal at high speed.After down-sampling from the first-time sampling point,the second-time sampling point was transmitted to the monitoring center on the ground,and changepoint detection algorithm was used to determine whether there is any leakage.If there is no leakage,monitor will be continued.If there is a leakage,the monitoring center will send instructions to the substation to request the saved first-time sampling data.Thirdly,the characteristics of the off-line pipeline leak acoustic emission signal were analyzed.The wavelet-Kalman combined de-noising method was used for the first-time sampling data of the leakage signal,and the pseudo cross correlation method was used to locate the leakage points with the filtered signal.Fourthly,according to the structure of underground branch pipeline,the experimental platform of branch pipeline is set up.The leak detection and location algorithm were tested in the whole system.The experiment showed that the method had good leakage detection and locating performance,and the rate of leakage detection could reach to 95%,and the locating accuracy was within 3.00cm.The theoretical method of leakage monitoring in this paper has high accuracy and reliability in the leakage detection of branch pipes,which is of great theoretical and practical value to take timely measures for pipeline leakage,reduce the economic loss and ensure the safety of the mines production.