Research And Application Of Micro Leakage Detection Method For Long Oil Pipeline Summary

As one of the five major transportation methods,pipeline transportation has become the preferred method of crude oil and refined oil transportation with economical,practical,safe and reliable characteristics.In China,nearly 80% of crude oil is transported through pipelines.China has formed crude oil that is spread all over the country and connects to foreign countries.The pipeline network has greatly promoted economic development.However,economy and danger often coexist.Due to the flammable,explosive,high temperature and high pressure characteristics of the transmission medium,once the oil pipeline leaks may cause catastrophic consequences,it must be discovered early and dealt with in time.This paper studies the problem of small leaks in long oil pipelines,which can avoid the expansion of leaks.Therefore,it has very important practical significance.The main research work is as follows:This paper presents a method for detecting small leaks in pipelines based on negative pressure waves.According to past experience,leak detection methods based on negative pressure waves are susceptible to the adjustment of operating conditions and have a problem of high false alarm rates.Therefore,this paper uses negative pressure On the basis of wave,after understanding,analyzing and comparing the signal denoising methods such as empirical mode decomposition,variational mode decomposition,local mean decomposition,and wavelet transform,we decided to use the combination of local mean decomposition and wavelet transform.The method is used to denoise,make full use of the advantages and disadvantages of the two methods,extract the eigenvalues of pipeline leaks,use support vector machines to identify the micro-leakage conditions,determine the occurrence of micro-leakage in the pipeline,and locate the location of the occurrence.After the experiment on the 2800 m circular pipeline,the small leak detection and positioning method of the long oil pipeline proposed in this paper can successfully identify the small leaks in the pipeline,the detection accuracy can reach 82%,and the positioning error is about 120 meters.