Ship Diesel Engine Condition Monitoring And Fault Diagnosis Based On Data Mining

With the rapid development of science and technology, the structure of the modern large equipment is increasingly complicated. With the equipment running continuously, it will be inevitably appear some various faults and cause incalculable economic losses and casualties.Fault diagnosis technology gives very important means of ensuring the equipment running safely, reliably and efficiently. It is of great significance in the aspects of reducing the equipment fault rate, prolonging the service life, reducing the economic loss and ensuring the safety of personnel.This paper applied data mining technology in ship diesel engine fault diagnosis, used support vector machine (SVM) algorithm and association rule mining algorithm, built a ship diesel engine condition monitoring and fault diagnosis simulation system which is based on data mining technology. This system realized the real-time monitoring of running status of the each ship diesel engine subsystems. It could also mining out the association relationship between the fault features by analyzing the ship diesel engine fault feature data. This system realized the intelligent fault diagnosis and assistant decision analysis. This article mainly launches the research from the following several aspects.Firstly, on the basis of reading a large number of relevant literatures, this paper analyzed and summarized the research status of the data mining technology and fault diagnosis technology in domestic and international. This paper divided ship diesel engine system into four parts according its basic structure and fault features, the fuel system, the lubrication system, the intake and exhaust system and the cooling system, and then analyzed the fault features for the each subsystem respectively.Secondly, this paper applied support vector machine algorithm for each of the ship dieselengine subsystems fault diagnosis. Discussed the basic principle of support vector machine algorithm, selected several characteristic parameters for each subsystem as experimental data,and simulated those data values to train support vector machine classification model, got the each ship diesel engine subsystem fault features classification model.Thirdly, this paper applied association rule mining algorithm for ship diesel engine whole system fault features, discussed the basic principle of association rule mining algorithm,analyzed the association relationship base on the historical fault transaction data and the fault classification data which is obtained from the SVM classification model, found out the implicit association relationship in the whole system fault information.Finally, this article designed a ship diesel engine condition monitoring and fault diagnosis simulation system and the human-computer interaction interface by using MATLAB GUI, used the MATLAB language to complete the code writing of association rule mining algorithm, and used the LIBSVM software packages to complete the training of the SVM classification model, used KEPSeverEx V4.0 (a kind of OPC server) to support the lower machine data acquisition and communicate with MATLAB, and ultimately implemented the ship diesel engine running state of real-time monitoring and fault diagnosis.