Research On Detection Methods Of Pressure Gas Leakage Sources

The petrochemical industry is one of my country’s pillar industries,which is of great strategic significance to the country’s economic development and safety engineering,and it is also closely related to human life.However,chemical companies are bound to bring safety issues in the production process,especially flammable,explosive and toxic gases.Once leaked,fires or even explosions will occur,which will bring disastrous consequences to people’s lives,national property and the ecological environment.Therefore,in addition to standard operations,the use of electronic technology to quickly and accurately detect is now a means.Among them,the sound source localization method has unique advantages for pressure gas detection.Research on the related technology of this method can provide a new way for the maturity of the technology in this field,which has important practical significance in engineering applications.For this reason,this subject proposes a pressure gas detection method based on the ultrasonic angle.This paper takes the research of the ultrasonic detection method of pressure gas leakage as the subject of research,and has done the following three aspects:(1)The noise reduction pre-processing of the leaked acoustic signal is the prerequisite for the detection and identification of the gas leak target.Therefore,this topic combines the wavelet packet analysis(WPA)and the variational mode decomposition(VMD)method.Extract the effective IMF component to reconstruct the signal,so as to achieve the purpose of removing noise and retaining the effective sound details of the target.Compared with the traditional filtering and noise reduction methods,the method proposed in this paper can effectively eliminate the noise in the signal,extract key information and has a higher noise reduction performance.(2)Before locating the gas leak source,the state of the pressurized gas needs to be monitored.This subject proposes a joint detection method based on IWOA-VMD and SVM.The WOA algorithm is improved from four aspects,the improved algorithm is applied to VMD,and the characteristics that can be used to characterize the leakage signal are extracted,and the gas leakage conditions are identified and classified.Experimental analysis proves that the method of this subject can effectively complete the monitoring work,with an accuracy rate of99.18% for judging leakage and a classification accuracy rate of 94.68% for the degree of leakage,which is superior to other methods in terms of time consumption and is more suitable for practical engineering applications.(3)Accurately estimating the location of the leak source is a key research issue in leak detection.When there is a leak,it is necessary to further identify and locate the abnormal situation so that the staff can go to the accident site in time and accurately to prevent the accident from further expanding.In this topic,a new DOA method for gas leakage sources is proposed in combination with the IWOA algorithm.The IWOA algorithm is introduced to search for DOA peaks.Compared with the traditional algorithm,the positioning accuracy is improved,the computational complexity is reduced,and the real-time performance of the algorithm is enhanced.Reduce the difficulty of engineering application.In terms of gas leakage positioning,the error of gas leakage location is small,the error of direction Angle is about 0.5°,and the error of pitch Angle is about 0.5°,which shows high stability and good estimation effect.