Research On Leakage Detection Of Long-distance Iron Concentrate Pipelines

With the wide application of pipeline transportation,the safe operation of pipeline has been paid more and more attention.As an important part of pipeline safety operation,pipeline leak detection has become an important research content in pipeline transportation.Pipeline leakage occurs will have very serious consequences,including economic and production losses and waste of natural resources and environment destruction,especially after the two can not be compensated,urgently need to be able to accurately and timely detecting system of pipeline leakage.Many problems need to be studied by the method of experiment.The design of slurry pipeline leak detection system provides experimental data base for pipeline leak detection.Because of the noise interference in the pipeline pressure signal,it is necessary to preprocess the pipeline pressure signal,which provides a good foundation for the next leak detection.In the actual industrial operation,the working condition of pipeline slurry is adjusted according to the actual production demand.At this time,the pressure signal of the pipeline is similar to that of the pressure signal in the pipeline leakage,which can easily lead to the interruption of pipeline transportation.It is of great significance to accurately detect the leakage of pipeline and avoid false report or omission.The main research work of this paper is as follows:(1)The paper introduces the experiment system of slurry pipeline leakage.According to the actual situation of the pipeline,selection of pipe material and wall thickness,the power system,the mixing and cleaning system and the measurement and acquisition system are designed.(2)Considering the problems of great difficulties in resolving pipeline pressure leak signal de-noising,a new method which is based on Singular Value Decomposition(SVD)was proposed.First,build Hankel matrix in original signal.Then do SVD decomposition of the selected signals.The component signals obtained by decomposition are used to finding out the sensitive component by using the sensitive factor.At last,the corresponding singular values of the sensitive components are reconstructed by the located factor.The experimental results show that this method,used as a pretreatment of signals,effectively removes the noise of pressure factor of mineral pipeline,which provides a good foundation for pipeline leak detection and location.(3)In this paper,a new method of pipeline leak detection is presented,which is based on the combination of wavelet decomposition approximate entropy and extreme learning machine.The pipeline pressure signal wavelet decomposition,selection of the main features of the first 3 layers containing components,approximate entropy and kurtosis value as the characteristic vector before the 3 layer of the component,the classification of feature vector.Based on the combination of wavelet transform approximate entropy and extreme leaRning machine,the pressure signal can be identified accurately and effectively.Through the research of long-distance pipeline leakage detection,it can ensure the safety of long-distance pipeline transportation system and reduce the loss caused by pipeline leakage.The design of slurry pipeline leak detection system provides experimental data for pipeline leak detection.lt provides a good foundation for pipeline leak detection by pre processing noise data.Pipeline leakage detection is based on the pre processed pressure signal.The simulation results show that the method can effectively and accurately identify the pipeline leakage.