Intelligent Detection And Localization Method Of Gas Pipeline Leakage

As one of the five major transportation industry which contain the highway,railway,water transportation and civil aviation,pipeline transportation plays a particularly important role in the development of national economy because it has the advantages of environmental,fast,economy.However,due to varieties of reasons,such as pipeline corrosion,natural damage,man-made destruction and pipeline failures,pipeline leakage often takes place.It not only caused a great deal of economic loss and environmental pollution,it may bring serious personal casualty accidents.Therefore,pipeline leak detection and location technology has important research significance to maintain the safe operation of the pipeline and protect people's lives and property safety.Acoustic method has been proved to have chacracteristic of low false alarm rate and high sensitivity.It is a development potential method which is used for gas pipeline leak detection and location.The acoustic method leak detection technology has many advantages,but the compressibility of the gas increased the difficulty of gas leak detection,the accuracy of the existing acoustic leak detection system is not perfect and very expensive,it also has the bottleneck in signal filtering,so it is not widely applied in the west-east gas pipeline in China.Acoustic method has two parameters,one is the propagation velocity in the pipeline,the other is the pipeline length.In practice,due to the complexity of pipeline surroundings,the state of disrepair and lack of construction drawings,these two parameters aren't very accurate,resulting wrong location detection in pipeline.Therefore,in this dissertation I study the propagation characteristics of leakage acoustic signal in the pipe,and expound the basic principle of blind system identification.Because the leak acoustic channels system has an impulse response with quite a long sequence,we use the cross-correlation fitting technique to establish the cost functions.Through the blind channel identification we can calculate the pipeline length and propagation velocity and analyze the parameters of acoustic method.The acoustic signals gathered by sensors are processed by a de-noising method using wavelet transform.And the method to choose wavelet functions,decomposition scales,threshold selection rules and threshold functions is made to determine the parameters of wavelet transform.Then we apply BP neural network and RBF neural network respectively to the leakage location.The simulation contrast and research on the signal attenuation rule verify the wavelet denoising method which means the positioning accuracy of RBF neural network is superior to BP neural network.The leakage acoustic signal has non-stationary characteristics,so this dissertation extracts the variation characteristics of acoustic signal amplitude and frequency by using local mean decomposition analysis method.The simulation shows that this method can effectively grasp the characteristics of different aperture leakage signal,has made the important basis for accurate judgment leakage aperture scale,establishing a rapid leak rescue plan and evaluate spill scale.Ulike liquid pipeline,the leakage signal flow characteristics in gas pipeline has great influence in pipe pressure,velocity,temperature and density.In view of the non isothermal gas pipeline,condition will be more complex.According to the changes of temperature and flow characteristics of gas pipeline,in my dissertation,I use the isothermal method and variable step precise Runge-Kutta method,which can be respectively used in isothermal and non-isothermal gas pipeline leakage detection and location.It has been verified by simulation that the effect of variable step precise Runge-Kutta method is better than isothermal method.