The Leakage Identification and Location of Ship Pipeline System Based on Vibration Signal Processing

The leakage of the ship’s pipeline system will bring great risks to engine equipment and seriously threaten vitality ship. In this paper, detection localization research are carried out based on vibration signal generated by leakage. First, finite element model is constructed obtain variation law when leaks out. Second, processed variational mode decomposition (VMD) radial basis function (RBF) neural networks. wavelet packet threshold noise reduction conducted before improve signal-to-noise ratio. Then, denoised decomposed VMD. effective component identified analyzing correlation coefficient between signal. center frequency energy used as feature vector train RBF network identify locate Finally, a test platform built under laboratory conditions. After processing data samples collected from test, trained leaks. sample identification results show that leak method VMD-RBF has high accuracy.