This paper used the acceleration sensor and dynamic signal acquisition instrument during the course of ship sailing to collect,analyze and process the real-time sailing data of marine diesel engine,judge the problems that will occur in the marine power plant and solve them in time.The fault diagnosis system of the marine diesel en-gines used the dimensionality reduction technology of principal component analysis(PCA)to preprocess the data,analyze them with K-means clustering method after the dimensionality reduction,and input the clustered data into radial basis function(RBF)neural network for training to obtain detailed information of fault type,which improved the efficiency and accuracy of fault diagnosis.
radial basis function neural networkprincipal component analysisK-means clustering
NETL
NSTL
万方数据
