Research On Leakage Detection Method Of Tailings Pipeline Based On Infrasound Wave And Negative Pressure Wave

Pipeline transportation has unique advantages in the transportation of resources such as oil,natural gas,and tailings,but due to human factors and pipeline materials,leaks are inevitable.Subsonic and negative pressure wave detection technologies are powerful means of detecting pipeline leaks and are widely used in the field of pipeline leak detection.Therefore,this article is based on the subsonic and negative pressure wave detection method for tailings pipeline leaks,and the main research work of the paper includes:(1)Research on signal analysis method of tailings pipeline leakage.The principle of subsonic and negative pressure wave detection for tailings pipeline was introduced,and the time-domain and frequency-domain characteristic values were described respectively.Shannon’s information entropy and complex entropy were introduced,and the method of signal extraction was studied.(2)Design of tailings pipeline leakage detection system.The overall scheme of the entire system was introduced in detail.The system uses subsonic and negative pressure wave sensors for leakage detection,and signal processing is carried out through data processing and remote terminal units(RTUs).The leak detection system is managed and monitored by the leak detection system base station software and the leak detection system center dispatching software.Finally,on-site installation and deployment of the hardware and software were carried out.(3)Research on denoising method of subsonic and negative pressure wave signals.This article proposes a new denoising method-improved wavelet threshold-CEEMDAN algorithm denoising.The algorithm uses the CEEMDAN algorithm to decompose the original signal,and then uses the correlation coefficient method to select IMF components.The improved wavelet threshold is applied to the 5-layer wavelet decomposition containing noisy IMF components,and the baseline drift frequency threshold is set to 1.5.Finally,the denoised IMF components and the remaining components are reconstructed.Through this research,it was found that this denoising method has a higher signal-to-noise ratio and waveform similarity coefficient,and the root mean square error is smaller.Compared with denoising methods based on VMD and improved wavelet thresholds,this method can more effectively suppress noise and retain the effective features of the original signal.(4)Research on pipeline leak detection methods.A leak detection method based on power spectral entropy and singular spectral entropy is proposed to detect tailings pipeline leaks.Firstly,the signal sample data under the non-leakage state of the pipeline is processed and the threshold range of power spectral entropy and singular spectral entropy is set.After noise reduction of the signal to be detected,if the power spectral entropy or singular spectral entropy of the infrasound and negative pressure wave signals obtained by using the power spectral entropy analysis method and singular spectral entropy analysis method are not within the set threshold range,the pipeline is judged to be leaking and the system is alarmed.In addition,signal time-frequency analysis was carried out for the pipeline under leakage and non-leakage conditions,and the Hilbert spectrum and marginal spectrum of the infrasound and negative pressure waves were analyzed when there were leaks or not.However,changes could not be observed from the Hilbert spectrum and marginal spectrum.Through experiments,it has been proved that the leak detection method based on power spectral entropy and singular spectral entropy can accurately detect leakage signals with a hole size of 20 mm or 25 mm and give an alarm,which has strong practicability and reliability and meets the technical requirements of the engineering project.