Research And Application Of Negative Pressure Wave Pipeline Leakage Detection Based On Combined Signal Processing

With the rapid development of social economy,the domestic demand for oil and gas is increasing year by year.As a convenient,fast and low-cost transportation mode,pipeline transportation is gradually paid attention to by the society.However,in recent years,with the increase of the proportion of pipeline transportation and the growth of the service life of the pipeline,it is inevitable to appear aging,corrosion and other phenomena,resulting in frequent leakage accidents of domestic pipeline,not only caused significant economic losses and casualties,but also caused a great impact on the environment.In order to discover pipeline leakage in time and further reduce the loss caused by accidents,it is necessary to implement effective leakage detection means.Therefore,it is very necessary to carry out the research of pipeline leakage detection and improve the accuracy of leakage location.Pipeline leakage detection methods based on pressure signals have been widely studied and applied in recent years.The negative pressure waves generated by leakage contain leakage location information,and effective leakage location can be realized by accurately calculating the time difference of negative pressure wave signals at both ends of the detected pipeline.However,the calculation accuracy of the time difference mainly depends on the extraction effect of the inflection point of the negative pressure wave.The noise mixed in the collection and transmission process of the pressure signal and the effectiveness of the inflection point extraction algorithm will affect the extraction result of the inflection point.It can be said that accurate extraction of singular value inflection point directly determines the accuracy of leakage location.Therefore,in view of these problems,this paper adopts the combined signal processing method to effectively and quickly extract the pressure signal of leaking pipeline and carry out research on the arrival of the inflection point of negative pressure wave signal.The main content is as follows:(1)Theoretical research on negative pressure wave method.According to the formation mechanism of the negative pressure wave,the mathematical model of the generation and propagation of the negative pressure wave is derived.Based on the model and the principle of the leakage detection and positioning of the negative pressure wave,the key factors affecting the accuracy of positioning are analyzed,and the velocity correction model of the negative pressure wave is established to improve the positioning formula.(2)Related research of wavelet transform method.As the selection of wavelet generating function and wavelet decomposition scale plays an important role in the whole wavelet transform method,it directly determines the applicability and limitation of wavelet transform.Therefore,for the pressure signal under the same measured leakage condition(leakage point 210 m,leakage amount 20%,length of detected pipe 771m),the signal-to-noise ratio(SNR)and root mean square error(RMSE)are taken as the evaluation indexes of noise reduction technology.Three types of wavelet systems commonly used in digital signal processing are selected:Symlets wavelet system,Daubechies wavelet system and Coiflets wavelet system,select the best wavelet generating function and wavelet decomposition scale.In this way,the abrupt part and noise in the signal can be distinguished effectively,the system noise can be filtered out,the signal-to-noise ratio can be improved,and the leakage judgment and location can be facilitated.(3)Research on combined signal processing technology.The characteristics of negative pressure wave signal and microseismic P confidence signal are analyzed,and two kinds of P-phase recognition algorithms in microseismic data processing,STA/LTA method and maximum kurtosis algorithm,are combined.Combining the advantages of each algorithm,an improved maximum kurtosis algorithm(W-S/L-MK method)based on discrete wavelet transform(DWT)and short-time average to long-time average(STA/LTA)is proposed to effectively and quickly extract the arrival of the inflection point of negative pressure wave signal.In this method,the denoised signal is obtained by discrete wavelet transform(DWT)decomposition,and the high-frequency noise part is discarded directly,and only the low-frequency detailed signal is analyzed.Then,STA/LTA method is used to initially screen the inflexion point of negative pressure wave and select the local kurtosis interval.Finally,the maximum kurtosis algorithm is used to calculate a series of kurtosis values in the local kurtosis interval,and the optimal singular value points are extracted according to the maximum kurtosis values.Experimental results show that the proposed method can combine the outstanding ability of retrieving the arrival information of the singular value of the pressure signal from the high resolution wavelet coefficients with the inherent reliability of obtaining the arrival time using the STA/LTA pickup method.The minimum relative error of leakage location is 1.47%.(4)Design and implementation of algorithm positioning software.In order to facilitate the application of the proposed algorithm in practical engineering,a negative pressure wave leakage detection and positioning software based on C language and MATLAB language is developed by using MATLAB.The software is easy to operate,including login module,home page module,leakage detection module,leakage location module and method comparison module,realizing the full set of functions of the proposed leakage detection and positioning method.