Leakage Monitoring Algorithms And Software Design Of Coal Mine Fluid Pipeline Network Based On Multi-sensor Fusion

Coal mine underground compressed air supply pipeline plays a key role in coal mine safety production.The pipeline is easily corroded and leaked due to the harsh natural environment of the coal mine,such as high air humidity.If we can monitor the pipeline in real time and determine the location of the leak source,it is of great significance to reduce resource waste and ensure safe production of coal mines.In this thesis,based on the detection of coal mine downhole pressure and water supply pipeline,the leakage detection and location of coal mine fluid pipeline network is studied.Considering the single sensor for detecting inaccurate result,while the multi-sensor data fusion can be well integrated variety of methods to improve the detection accuracy,so this technique will be applied to the multi-sensor fusion fluid pipeline leak detection,and the leakage point location algorithm conduct in-depth research.The main research work and innovation are as follows:(1)Research on leakage detection algorithm of coal mine fluid network leakage detection.The wavelet threshold denoising algorithm was used to process the acoustic emission signal and the pressure signal.Because in some cases the sensor is far away from the leak point,the time-domain features of the acoustic emission signal collects are not obvious and difficult to extract.Therefore,when extracted the feature vector of the acoustic emission signal,the author used an algorithm that combines the synchronous wavelet transform(SWT)with Non-negative matrix factorization(NMF).And for the extraction of basic time domain eigenvectors that are related to the pressure signal,The eigenvectors of the two signals are respectively input into the support vector machine for discriminating,and the respective output results are combined with the D-S evidence theory multi-sensor technology for fusion judgment,Through the above method,the pipeline running state can be finally determined.(2)Research on leak location algorithm of coal mine fluid network.Used generalized cross-correlation and correlation optimal weighting functions to calculate the time delay estimation of acoustic emission signals.And by comparing the characteristics of different weighting functions,the PHAT weighting function was selected to obtain the time difference of the signal reaching the adjacent sensor.Then the researcher can get the specific location where the leak occurred.(3)Using Lab VIEW as the software design platform,Combined with the coal mine fluid network leakage detection and location algorithm,a set of functional software systems were designed that mainly including real-time signal display module,signal denoising processing module,pipeline leakage detection and positioning module,data storage module,etc.(4)A piping system model was established based on the underground pipeline structure to simulate pipeline leakage.The thesis actually tested the detection,localization algorithms and Software throughout the system and analyzed the test data.The experimental results showed that the system had good performance,and the rate of leakage detection could reach to 98.1%,and the average location error could reach to 4.9cm,and could meet the leakage detection and positioning requirements of coal mine fluid pipeline network.