| The fault diagnosis and state recognition of marine diesel engine is an important research direction of marine diesel engines,and it is a research hotspot in the field of ship and automation at home and abroad.Information fusion technology has played a huge role in various fields such as national life and military,and its application in the field of ship diesel engine fault diagnosis is still in an immature stage.In the fault diagnosis based on information fusion,there is a lot of information available,but only by selecting and using effective information to analyze the equipment can improve the accuracy of fault diagnosis.Researching and exploring more practical methods of fault diagnosis applied to marine diesel engines will help to improve the safety of navigation and operation of ships.In this paper,the AVL BOOST simulation software is used to establish the mathematical model of the MAN 8L51/60DF diesel engine working process,and the reliability of the model is verified by the diesel engine bench test data.The results show that the calculation error is within 3%.In the diesel engine simulation model,its normal working condition,excessive oil supply in a single cylinder,excessive fuel injection advance angle,too small fuel injection advance angle,premature exhaust valve closing,and exhaust valve closing are respectively carried out.under rated conditions.A total of 96 sets of data were calculated for the training and verification of the subsequent fault diagnosis model.In order to reflect the operating state of the diesel engine from many aspects,two BP neural networks are designed to diagnose the diesel engine respectively with different training functions.However,due to the limitations and shortcomings of the traditional BP network,the accuracy of single neural networks in fault diagnosis is low.In order to further improve the accuracy of fault diagnosis,a fault diagnosis method based on Elman and GRNN neural network is used,and a fault diagnosis model based on multi-neural network and D-S evidence theory is designed.When the information fusion is carried out,the diagnosis result of the neural network is used as the basis for the distribution of the basic credibility,which avoids the defects caused by the subjectivity of the human.Finally,the fault data simulated by AVL BOOST is used to test the neural network diagnosis model and the diagnostic model combined with the neural network and DS evidence.The experimental results show that the fault diagnosis results of each model are basically consistent with the actual state of the diesel engine,but the accuracy and confidence of the model based on information fusion are significantly improved,Compared with the neural network,the confidence and accuracy of the model based on information fusion are significantly improved,and the effectiveness of the method in diesel engine fault diagnosis is fully verified. |
