Research On VMD Algorithm In Pipeline Leak Detection

The pipeline leak detection technology is the most important means to ensure pipeline safety.In the maintenance operation,the pipeline needs to monitor the pipeline signal in real time,accurately determine the pipeline operation,timely diagnosis of the existence of leakage hazards,to avoid the occurrence of security incidents.The infrasonic wave signal produced from natural gas pipeline leakage belongs to non-stationary signal,and is seriously interfered during its spread in the pipeline.These lead to high level of misjudgment rate of leakage and misstatement,missing report and wrong report.Self-adaptation time-frequency analysis method is an important method of analyzing non-stationary signal,targeting at the features of infrasonic wave produced during the laboratory pipeline leakage,in this paper,the Variational Mode Decomposition(VMD)algorithm based on the correlation coefficient is used to select the default dimension K,and then the pipeline feature extraction is carried by VMD algorithm based on mutual information,so as to to distinguish the identification of different working conditions in the pipeline,therefore improve the system's judgment accuracy of the pipeline leakage.The main work of this text includes:1.Look up to relevant thesis on pipeline leakage detection,have a deep understanding of relevant methods of pipeline leakage detection,choose VMD methods and apply them into the leakage detection of natural gas pipelines;analyze the malfunction existed in the pipeline leakage detection,classify the malfunctions,and finally ascertain the target type of the mode identification;finish the signal collection under different pipeline operations.2.Targeting at the fact that VMD algorithm has the issue of difficulty in selecting default dimension K and effective IMF components after resolving,it proposed a VMD method which is based on relevant parameters to optimize the parameters,and use it in detecting pipeline leakage signals.Firstly,verify the effectiveness of the VMD algorithm which is based on relevant parameters via simulation signals;then apply the VMD algorithm based on relevant parameters to the detection of pipeline leakage signals to verify the viability of the algorithm.3.Aiming at the problem that the high frequency part of the decomposition of VMD is affected by noise interference,the VMD method based on mutual information is proposed.Firstly,the validity of the VMD algorithm based on mutual information is verified by thesimulation signal.Finally,the VMD algorithm based on mutual information is applied to the pipeline leakage signal detection,and the feasibility of the algorithm is verified.Compare the VMD algorithm to verify the effectiveness of this method.4.For different signal conditions pipeline,the VMD algorithm based on the correlation coefficient is used to select the default dimension K,and then the pipeline feature extraction is carried by VMD algorithm based on mutual information,so as to to distinguish the identification of different working conditions in the pipeline,The accuracy and feasibility of the method are verified by the experimental results.