Key Technologies Research Of Pipeline Leak Detection In Strong Noise Environment

The pipelines for fluid transportation are the most convenient and economical means of transportation.It is also the main transportation method for important energy sources such as oil and natural gas.Therefore,the safety of pipeline is also very important,which is valued by the pipeline management department,which also directly promotes the continuous development of pipeline leak detection technology.Pipeline leak detection technology involves the integration of sensor technology,signal processing technology,automatic control technology and many other fields,and the methods of pipeline leak detection vary widely from the initial manual inspection line to the present.Remote real-time monitoring,from handheld inspection equipment to artificial intelligence and big data processing.Pipeline leak detection sensors are generally installed in the booster station at both ends of the pipe.The booster station uses a fluid drive device（compression pump or motor）to apply a driving force to the pipe fluid to allow fluid to flow in the pipe.It is interfered by the noise of the pipeline fluid drive device,and the frequency of this noise is mixed with the frequency of the pipeline leakage noise.It is difficult to filter out by ordinary filtering methods.Therefore,the denoising technique for leaking signals under strong noise environment is also the focus of the pipeline leak detection technology.Firstly,the self-mixing interference（SMI）model is analyzed according to the three-mirror cavity（FP）model,and the phase equation and power equation of SMI are derived.Theoretically,the self-mixing interference signal has the same fringe resolution as the traditional two-beam interference signal.This paper also studies the effects of optical feedback factors,external object vibration parameters and multiple feedback on self-mixing interference signals,and analyzes fringe counting method,phase unwinding algorithm,main frequency order algorithm and phase modulation self-mixing interference signal reconstruction technology.Secondly,based on the noise spectrum aliasing between the pipeline leakage signal and the fluid drive device,a pipeline vibration measurement technique based on self-mixing interference is proposed.The jump point detection algorithm based on local maximum is proposed.Then it is applied to the phase unwinding algorithm for self-mixing vibration detection.The self-mixing interference vibration sensor experimental measurement system is built to realize the self-mixing interference vibration reconstruction.The simulation and experiment were carried out.The self-mixing interference vibration sensor was applied to the pipeline vibration detection.The experimental results were verified from three directions:effectiveness,linearity and frequency response.The experimental results show that the SMI vibration sensor has similar measurement accuracy to the laser Doppler vibrometer in pipeline vibration detection.Then,aiming at the selection of the total number of modals（represented by K）and the determination of effective IMFs,a signal denoising algorithm based on VMD and mutual information is proposed.The original signal is decomposed into multiple Effective eigenmodes（IMFs）by VMD.According to the difference between the original signal and the summed sum of the IMFs,the signal filtering is realized by reconstructing the signal.The algorithm can effectively realize the fuzzy optimization of K and solve the problem of K selection in the VMD algorithm.The experimental results show that the proposed method can achieve the fuzzy optimization of K and retain the important characteristics of the original signal.By evaluating the correlation coefficient,MSE andSNR_{ou t},In the filtering of pipeline acoustic wave sensor and self-mixing interference sensor,VMD has better denoising performance than filtering methods such as EMD and wavelet transform.Finally,for the noise aliasing of pipeline fluid drive device and pipeline leakage signal under strong noise environment,the pipeline leakage signal detection algorithm under strong noise environment is proposed based on Joint denoising algorithm.The acoustic wave sensor is used to detect the pipeline gas leakage signal,and the self-mixing interference sensor is used to detect the vibration noise signal of the outer wall of the pipeline caused by the pipeline compressor.The Joint denoising algorithm is used to denoise the two signals to obtain the pipeline leakage signal.The leakage of the natural gas pipeline was simulated through the laboratory pipeline to verify the effectiveness of the pipeline leak detection technology in the strong noise environment proposed in this paper.This paper simulates five scenarios in which a pipeline leaks under strong noise conditions.The experimental results show that under five kinds of strong noise conditions,the denoising effect of the Joint denoising algorithm is 0.23-1.72dB higher than the spectral subtraction algorithm.