Research On The Enhancement Method Of Weak Leakage Signal Of Oil Transmission Pipeline

Due to the adjustment of pipeline transportation processes,changes in operating conditions,the existence of pipeline coating layer,soil cover for buried pipelines,and the attenuation of leakage signals as they propagate through the medium,serious signal attenuation is often observed in leakage signals detected upstream and downstream of pipelines.This is reflected in the pressure drop on the negative pressure wave,which exhibits a small pressure drop amplitude and a slow pressure drop rate.Existing signal filtering methods are also unable to effectively enhance such weak leakage signals,leading to the occurrence of false alarms.Therefore,the study of signal enhancement for weak leakage signals has important theoretical significance and practical application value,as it can effectively reduce the occurrence of false alarms.In view of the fact that the signal-to-noise ratio(SNR)cannot truly reflect the optimal filtering scale,this paper proposes a multi-scale representation method for the amplitude of the signal,which is used to reliably characterize the prominence of sub-signals at different filtering scales.This method uses the peak value of interval sub-signals as a feature,and obtains a single-scale representation of the amplitude of each interval sub-signal under fixed singleconnection distance threshold conditions through an abnormal sub-signal detection method based on hierarchical clustering.By accumulating the single-scale representation values of sub-signal amplitude under different single-connection distance threshold conditions,a multi-scale representation of the amplitude of each interval sub-signal is obtained.The multi-scale representation of sub-signal amplitude gives a relative perspective on the prominence of its amplitude compared to other sub-signals,which is independent of the noise power at different filtering scales and is only related to the filtering scale.By accumulating the multi-scale representation values of sub-signal amplitude under different filtering scales,a multi-scale representation function of a frame signal is obtained.The sub-signal corresponding to the negative extremum in the multi-scale representation function is taken as the target sub-signal to evaluate the signal enhancement effect.By comparing the size of the negative extremum of the corresponding position of the target sub-signal under different filtering scales,the optimal filtering parameters can be selected.Then,the optimal filtering result is multiplied by the multi-scale representation function to actively suppress the background noise.Considering that multiple abnormal sub-signals may occur in a frame signal,a correlation test is performed on the upstream and downstream target sub-signals to ensure their matching.A band-pass filter combining low-pass filtering and high-pass filtering is used as the technical means for signal enhancement in this paper,and the signal enhancement method based on multi-scale representation of signal amplitude is effectively verified.The low-pass filter uses multi-scale wavelet decomposition,and the high-pass filter uses a first-order RC high-pass filter that simulates quasi-static charge amplification(with fixed feedback resistance and feedback capacitance).The adjustment of its cut-off frequency is achieved by iteratively adjusting the wavelet decomposition scale and the discrete frequency of the high-pass filter transfer function.Comparison experiments with other methods show that the weak leakage signals enhanced by the signal enhancement method proposed in this paper have higher SNR,smoother background noise,and significant signal enhancement effects,and can cope with the interference of different operating conditions and background noise.The leakage detection and location results further demonstrate that this signal enhancement method can improve the detection accuracy and location precision of weak leakage signals.