Research On Marine Diesel Maintenance Decision

In order to reduce the influence of man-induced factors in the course of marine diesel maintenance decision and to normalize maintenance management, this paper puts forward a Union Fuzzy Neural Network model to decide the maintenance on the basis of Selective Maintenance System theory. Such results are concluded from Matlab simulation tests:1. Marine Diesel Union Fuzzy Neural Network model can not only reflect the fuzzy character and logicality of Main Engine, but also avoid the joints between irrelevant parameters, simplify the network structure, and decrease the calculation.2. A great amount of oil analysis data and performance parameter data were collected and trimmed into several samples in this paper. The results of Matlab programs show that this model is able to learn the law from history data perfectly and quickly. Using this FNN decision model, the marine maintenance manager can improve the scientificity of maintenance management, lessen the influence of man induced factors, and it has much practical value.3. This paper compares the union FNN model with the simple FNN and Fuzzy Synthesis Decision model, and get that the union FNN model can reach the conclusion accurately but the calculation is too much. While the simple FNN can also get the right results, it can't tell which part needs repair. The FSD is simpler but more accessible to subjectivity, and it can't learn from experience.The research of this paper provides a new decision method for marine diesel maintenance management, It is an attempt to scientize and to normalize the maintenance management.