Development Of Performance Database And Research On Neural Network Diagnosing For Marine Diesel Engine

With the rapid development of China's shipbuilding industry, the marine diesel engine manufacturing industry of China has also been developed fast, which results in the increasing of the number and the types of the diesel engines produced and the exponentially growing of the information. In the case, the previous recording method is difficult to deal with the exponentially growing data, and these problems can be successfully solved by the development of the computer and the database technology. According to the needs of the company informatization, the parts of the functions of the marine diesel engine digital management system (DMS) which based on Justep X3 software platform and SQL Server database are developed, a new diagnosis method based on neural network algorithm about several typical faults of the marine intelligent diesel engine is systematically researched. The main work is as follows:1) According to the requirement of users, the demands of the system are analyzed. The databases of the basic information management subsystem and the technical service support subsystem are designed.2) The design and development work of the basic information management subsystem and the technical service support subsystem is accomplished. According to the method of software testing, their functions are tested, which can verify the system's rationality and reliability.3) The work flow of the DMS system is designed and realized, which including the subsystem process of the basic information management, the subsystem process of the test management and the subsystem process of the technical service support.4) The feasibility of applying neural network algorithm to diesel engine fault diagnosis is discussed, and the method of the faults diagnosis for the marine intelligent diesel engine is researched. The diagnostic model based on the BP neural network for several typical faults of the intelligent diesel engine is built, and its structure and parameters are determined.5) The fault information database of the study object is established based on the simulation data and the test data, and the fault diagnosis subsystem is completed. The simulation data mainly is come from the AVL BOOST simulation model established by other researchers, and the experimental data is come from the test data of Hudong Heavy Machinery Company Ltd.