Fault prediction of ship machinery based on gray neural network model

Aimed at the limitation of feedforward and feedback ANN, and the shortcoming that the diagnostic characteristic parameters are considered separately in conventional fault forecast method for machinery equipment, the multivariable gray model, MGM(1,n), and RBF network are introduced into fault prediction, which allows characteristic parameters to be described from the viewpoint of systems. It predicts future characteristic parameters considering the past and current machinery information, then RBF network is used to predict online. The fault prediction example indicates that the model has good prediction precision. It offers an effective method for reliable real-time fault diagnosis.