Research On Oil Pipeline Detection And Location Based On Exponentially Decreasing Inertial Weight PSO

According to the advantages of high safety and fast transportation,pipeline transportation has become an important part of energy transportation.However,due to the harsh external environment of the pipeline,the friction and corrosion of energy transportation in the pipeline,as well as the illegal elements of stealing oil and other factors,the pipeline leakage accidents occur frequently,which not only wastes energy and pollutes the environment,but even endangers life safety.Therefore,the detection of pipeline leakage and the location of leakage point are extremely important.Acoustic signal has become a major research direction in pipeline leakage detection with its advantages of rich information,low attenuation rate and obvious changes after leakage.This paper uses MEMS acoustic vector sensor as the detection device,and proposes a detection and location method of oil pipeline based on exponentially falling inertial weight PSO.Firstly,this paper analyzes the acoustic signals of oil pipeline leakage,classifies the acoustic signals according to different sound media,and studies the propagation theory of acoustic signals in soil.The principle of leak location based on acoustic signal is studied.Secondly,a piezoelectric MEMS acoustic vector sensor is designed according to the characteristics of the leakage signal and its propagation in the soil,and the vibration model,frequency characteristics and sensitivity characteristics of the structure are analyzed theoretically.The finite element software was used to simulate the shape and parameters of the sensor,and the influence law of the structural parameters on the sensitivity and frequency band of the sensor was obtained.The optimal parameters of the sensor were determined according to the characteristics of the leakage acoustic signals.Vibration experiments were carried out to test the performance of the sensor,and the sensitivity of the designed sensor reached 41 m V /g,and the frequency response range was 0-7200 Hz,which met the detection requirements.Thirdly,in order to solve the problem that MUSIC algorithm has poor bearing estimation effect when the SNR is low,this paper proposes a nonlinear exponential function optimization particle swarm optimization algorithm,and uses this particle swarm optimization algorithm to optimize MUSIC algorithm.Specifically,the powerful search ability of particle swarm optimization algorithm based on inertia exponential decline is used to optimize MUSIC algorithm and improve the orientation accuracy of MUSIC algorithm in azimuth estimation.According to the single vector sensor simulation experiment and linear sensor array simulation experiment,the directional capability of the single vector sensor designed in this paper and the proposed optimization algorithm are compared and analyzed.Besides,it is verified that the proposed MEMS acoustic vector sensor has good directional ability,and the accuracy of azimuth estimation is improved by the MUSIC algorithm optimized by PSO with the weight of inertia index decreasing.Finally,field simulation experiments of single vector sensor and linear sensor array are carried out in this paper to further verify that the MUSIC algorithm optimized by the inertia index weight reduction PSO proposed in this paper can reduce the relative error of azimuth estimation by about 3%.