Research On Hole Leakage Detection Method Of Atmospheric Vertical Cylinder Oil Storage Tank

The oil storage tank is an important equipment for oil storage.With the increase of its use time,the probability of failure increases.Among all the failures of the oil storage tank,leakage is one of the most dangerous failure types.At present,the research on leak detection of storage tanks is mainly focused on pressure storage tanks.There are few studies on leak detection of atmospheric storage tanks.The leakage characteristics of pressure storage tanks are obvious,while the leakage characteristics of atmospheric storage tanks are not obvious,which increases the difficulty of leak detection.Most of the oil stored in the oil storage tank is flammable and explosive.When the oil storage tank leaks,it is easy to cause fire and explosion accidents.Finding the oil storage tank leaks in time and effectively and locating the leak is helpful to realize rapid maintenance,which can greatly reduce the occurrence of oil storage tank accidents and avoid heavy loss of personnel and property.Therefore,it is of great significance to study the method of hole leakage detection for atmospheric vertical cylinder oil storage tank.This article mainly carries on the following several aspects research.Firstly,by referring to the pressure based leak detection methods in pipelines,a leak detection method of atmospheric vertical cylinder oil storage tank with pressure signal is proposed.In order to explore the change characteristics of the internal fluid pressure during the hole leakage of the atmospheric oil storage tank,the mathematical modeling of the hole leakage process of the atmospheric oil storage tank was carried out to clarify the relationship between the liquid level,water pressure and time in the hole leakage process of the oil storage tank.Fluent software is used to simulate the hole leakage process of oil storage tank.The simulation results conform to the hydrostatics,which shows the correctness of the mathematical model.At the same time,it is found that the water pressure at the leakage hole will change suddenly when the oil storage tank hole leaks.Secondly,in order to verify the sudden change of water pressure at the leakage hole in the simulation results,a simulation experimental platform for the hole leakage of the storage tank was built to collect the monitoring signals such as the liquid level,water pressure and flow rate during the leakage of the storage tank hole.In order to solve the problem that the monitoring signal is disturbed and fluctuates greatly,which affects the effect of data feature extraction and analysis,wavelet analysis is used for the monitoring signal de-noising.After the impulse noise is removed by the median filter of the signal,the heursure and soft threshold function are selected to use coif5,db6,and db5 wavelets to carry out 5-layer decomposition and reconstruction respectively,and the de-noising effect is better.Then wavelet decomposition and empirical mode decomposition are performed on the water pressure signal,and it is found that the water pressure signal does have a mutation.Finally,the leak detection method of oil tank hole based on machine learning is studied.The time domain and waveform features of water pressure signal after wavelet de-noising are extracted,and the support vector machine model is established to realize the detection of oil storage tank leakage.The water pressure signal was used to equalize the sample set,the leakage hole aperture identification models based on BP neural network and extreme learning machine were established respectively to realize the leakage hole aperture identification.The experimental results show that the methods described above are efficient and the extreme learning machine is faster to recognize than the BP neural network.,which is suitable for the identification of more sample data or multiple leak aperture of oil storage tank hole leakage.