Establishment of leakage detection model for oil and gas pipeline based on VMD-MD-1DCNN

Abstract In order to solve the problem of recognition error caused by noise interference in oil and gas pipeline signal traditional leakage detection relies on expert experience extract features manually, an model based deep learning is proposed this paper. The consists data preprocessing part pattern part. Firstly, a denoising algorithm variational mode decomposition (VMD) Manhattan distance (MD) reduce quality subsequent process. Secondly, combined with one-dimensional temporal characteristics signal, denoised VMD-MD used as input convolution neural network (1DCNN) learning, then are learned independently through structure. Finally, structure parameters optimized analyzed build optimal according experiment. experimental results show that, compared other existing models, ensemble VMD-MD-1DCNN has better improvement each evaluation index, which verifies effectiveness its application detection.