Research On Valve Inner Leakage Detection Technology Based On Convolutional Neural Network

With the increasing mileage of natural gas pipeline construction in China,valves play an indispensable role in the transportation of natural gas pipelines.If the valve leaks,it will bring economic and environmental damage,and pollution,fire and explosion accidents caused by valve inner leakage may occur.The laboratory's natural gas pipeline valve inner leakage detection device converts the acoustic emission signal into a voltage signal,manually extracts the characteristic parameters of the signal,and detects the inner leakage signal of the laboratory natural gas pipeline valve.Natural gas stations have a lot of noise,and it is difficult to detect leakage noise and environmental noise.The method of artificially extracting characteristic parameters cannot be used in the field.In this paper,the deep learning convolutional neural network is used to detect the leakage signal spectrum.The computer automatically extracts characteristic parameters to identify whether the valve in the noisy environment has leakage,it can eliminate the disturbance of changing the feature parameters manually selected due to the change of detection environment.The signals measured from the laboratory environment to the field are used to identify the leakage signal in the ambient noise in Matlab.At the same time,the Alex Net and VGG16 convolutional neural networks were used to detect the leak signal of the valve,and Alex Net was selected to identify the leak signal.In the hardware of the detection device,the detection environment is mostly a dangerous area such as a natural gas station.The leak detection device of the natural gas pipeline valve developed by the laboratory itself has no explosion-proof function.The probe of the device is attached to the front and back end of the valve under test,if the natural gas pipeline leakage or gas valve leakage occurs,there will be great safety hazards,This paper will carry out explosion-proof design of laboratory equipment,in line with national explosion-proof standards.