Research On Valve Leakage Detection Instrument Development And Internal Leakage Rate Quantitative Regression Analysis

Valve internal leakage is great risk faced by natural gas pipeline transmission.However,there is no specialized quantitative detection instrument for natural gas valves in domestic.Acoustic emission has been applied to the leakage detection in gas pipeline valves as an excellent nondestructive detection technology.In this thesis,the valve leakage detection system was developed based on the previous experimental results obtained from our laboratory.The main research contents are as follows:(1)To demonstrate the feasibility of valve internal leakage detection on the basis of acoustic emission technology,the acoustic emission detection technology and the valve internal leakage detection instruments were investigated.The reasons of valve internal leakage and the characteristics of jet noise were analyzed.Besides,different detection methods of valve internal leakage were compared.(2)To optimize the detector of valve internal detection,the hardware and software of the original instrument was remade.To make integrated design in the hardware,a new high-speed acquisition card and an embedded processor were adopted.Some new functions were added to the new software for the purpose of later updates and optimization,such as plug valve internal leakage,updating regression model and the Access database management.(3)Due to the complex environments of valve internal leakage detection,an improved method in wavelet threshold de-nosing was adopted to obtain pure internal leakage source and signals characteristic parameters.The Linear Discriminant Analysis(LDA)method was adopted to reduce dimensions of the signal parameters.Then,the main factors(kernel function,penalty factor C and kernel parameter g)were optimized and selected.Furthermore,the leakage rate regression model was established based on the support vector machine(SVM)theory.(4)The experimental platform of valve internal leakage detection was constructed in the laboratory.In order to satisfy the demand of the natural gas station,both the hardware and software of the acoustic detection system were examined and optimized based on a large number of laboratory and field tests.