Intelligent Diagnosis And Positioning Method Research Of Pipeline Leakage Based On Multi-sensor Information Fusion

With the rapid development of the economy, the people demand more and more on thewater, oil and other fluids, so the safe operation of the fluid pipeline has become a veryimportant task. So the studies of the fluid pipeline leak diagnosis and location havebecome very popular, in most cases, a single method result in inaccurate in test, however,multi-sensor information fusion technology effectively fuses variety methods. Therefore,according to the above idea, this paper put forward a method that is multi-sensorinformation fusion applied to the study of intelligent diagnosis and location of the fluidpipeline.The paper systematically studies the pipeline leak intelligent diagnosis and locationmethod, and the main research works are as follows:First, the paper introduces the significance of the fluid pipeline leak intelligentdiagnosis and localization studies, its objectives and trends. briefing introduction somemethods of pipeline leak detection, detailedly analyze the negative pressure wavedetection technology and sonic detection technology. Due to the presence of noise, wemake use of wavelet transform denoising for the date of negative pressure wave signal andsonic wave signal, and extract signal eigenvalue after denoising.Second, due to the lack of traditional single pipeline leak intelligent diagnostic methodthe paper puts forward a method that is based on multi-sensor information fusion, andintroduces detailedly multilayer feedforward neural network algorithm, and itsimprovement, adopting more effective the BP-LM optimization algorithms, the resultsshow that based on the neural network multi-sensor information fusion the pipeline leakintelligent detection conducive to improve the accuracy of the leakage fault detection.Finally, the ultimate goal of analysing pipeline detection leaks is to discover pipelineleak location. In order to solve this problem, we adopt negative pressure wave and sonicwave to locate, but these results are not accurate for single methods, so the informationfusion theory of D-S evidence is used to fuse the result of these two methods to locateleakage point, the results show that it has higher precision and more accurate location thansingle detection methods.