Prediction Of Ship Main Engine Thermodynamic Parameters Using RKGM-AR Model

The working condition of marine diesel engine,"heart" of the ship, affects not only the safety of navigation, but also the cost and profits of the shipping companies. In order to solve the problems of insufficient repair and over repair in the process of healthy maintenance and management of marine diesel engine, present repairing has been transferred from time-fixed repair to fault prediction and health management. Meanwhile, trend forecast and analysis of marine diesel engine thermodynamic parameters provides the technical support for this transition, thus realizing the fault prediction of diesel engineTaking No.1cylinder exhaust temperature of a training ship’s main engine as the main forecasting object, this paper proposes a combined forecasting model to conduct trend forecast and analysis, thus fulfilling the fault prediction of the diesel engine.Firstly, the regular thermodynamic parameters of diesel engine are analyzed. Afterwards, grey relational analysis method is theoretically elaborated and then used to carry out the clustering analysis of thermodynamic parameters, so as to obtain correlation parameters of exhaust temperature.Secondly, the paper analyzes the advantages and disadvantages of several common forecasting methods and proposes that the combined model is the future development trend. The combined forecasting model is established through the combination of the grey prediction model improved by fourth order runge kutta method and time series AR model meaning that advantages of the two models above are utilized.Then, through the study of calculation method for exhaust temperature alarm limit and classification of warning level, the warning function is realized. The exhaust temperature is taken as the main sequence. The average exhaust temperature, scavenge temperature, outlet temperature of main bearing lubricating oil and cylinder cooling water are regarded as subsidiary sequences. By selecting the sample data of the five parameters above when the diesel engine exhaust temperature experiences changes in steady and rising rate respectively, the combined prediction method is employed to predict and analyze the trend of exhaust gas temperature.Finally, combined forecasting model was applied to the real ship, and comparison and error analysis between the exhaust temperature prediction value and measured value are carried out to verify the validity of the prediction model.