Research of Marine Diesel Engine Condition Detecting Base on BP Neural Network and Spectrometric Analysis

The content of metal elements in the marine diesel engine lubricating oil is an important reflection of health condition of marine diesel engine friction components. By predicting the metal elements content in the oil, the mechanical faults in the diesel engine can be detected in advance, thus the severe damage in engine will be avoided. The running condition of marine diesel engine is complicated at sea, the content of lubricating oil is affected by many factors, the content trend of metal elements is hard to be effectively predicted in traditional methods. A new prediction method based on BP neural network is raised in this paper, meanwhile, the modeling and simulation are utilized with MATLAB. The prediction method is applied in the content of metal elements of some marine diesel engine lubricating oil, the relative error on average of results is 2.11%~2.72%, it can satisfy the demand of condition detecting in marine diesel engine.