Localization Algorithm Of Gas Leakage Source Based On Compressive Sensing

With the rapid development of modern petrochemical industry,many inflammable,explosive,toxic and harmful gases are produced during the production,use,transportation and storage,to frequent accidental leakage,casualties,conomic losses and environmental damage.Therefore,the techniques of gas leak detection and location have attracted many research attentions,especially the techniques of ultrasonic leak detection,which possess the strong penetration ability,high sensitivity,efficiency,accuracy as well as some other characteristics,and have became the research highlights in the recent years.For a long time,many domestic and foreign scholars have conducted many research studies on ultrasonic gas leakage detection,but the current theory mainly focuses on the method of gas leakage detection.The research of sound signal processing in gas leaking has been mostly disregarded in the existing studies.In addition,the classical methods of DOA suffer many issues during the process of acoustic leakage signal such as low resolution,poor anti-noise,which are not suitable for the applications of low snap,multiple leakage sources and fewer array elements.This paper first focuses on the research status of ultrasonic gas leak detection technology and signal DOA estimation method,and studies the array form of uniform linear array and the basic mathematical model of narrowband signal.Two common classic DOA estimation algorithms are introduced,namely CBF algorithm and MVDR algorithm,and simulation experiments and analysis of signal source positioning are performed on these two algorithms based on the MATLAB platform.The traditional signal acquisition method and the compressed sampling method are compared and analyzed,and the specific steps of signal reconstruction using the basic theory of compressed sensing are described.According to the compressed sampling process,the sparse representation of the spatial domain and the convex optimization signal reconstruction algorithm,a signal DOA estimation model and l₁-svd algorithm based on the compressed sensing theory are established.MATLAB simulation is conducted for comparing the performance of the proposed algorithm and the classical DOA estimation algorithm,and it is proved that the proposed algorithm possesses the advantages of saving the number of array channels,saving sampling data and better noise resistance.Furthermore,according to the mechanisms and characteristics of the ultrasonic wave that is generated by the gas leak hole,appropriate ultrasonic transducer and data acquisition card were selected,and a gas leak detection system based on compressed sensing was built based on Lab VIEW in order to detect the leakage and find the leak point.The experiment indicates that our system can accurately locate gas leakage sources within an effective detection distance,and the method can locate multiple leakage sources with low cost,high efficiency and high precision,which make it widely prospective to various applications.