Fault Diagnosis Method of Ship Hydraulic System Based on KPCA-PNN

The accuracy and rapidity of fault diagnosis of ship hydraulic system has always been one of the key areas of modern ship research. This article analyzes the failure mode of a certain type of ship hydraulic system, focusing on the hydraulic system's nonlinearity, non-Gaussian distribution and excessive dimensionality of collected data the problem. A fault diagnosis method (KPCA-PNN) based on the combination of kernel principal element and probabilistic neural network is proposed. Simulation results show that this method can detect and identify fault types more quickly and accurately than PNN and PCA-PNN methods.