面向动力需求的汽油机性能与油品组分关联性分析

In order to explore the factors and ways of affecting the dynamic performance of internal combustion engines, the engine performance tests under steady-state conditions were carried out based on a full-parameter controllable special engine test platform, and a correlation study of engine dynamic performance parameters based on BP neural network was proposed. Determine the typical combustion process parameters such as ignition advance angle, air-fuel ratio, combustion phase, and combustion duration as dependent variables in the middle layer, normalize the indicators, and establish an input layer based on BP neural network (engine control parameters, oil quality, etc.) -intermediate layer (combustion process parameters)-output layer (engine output power) correlation analysis model, the correlation coefficient is used to evaluate the correlation between each index parameter, which provides a reference for reverse fuel design oriented to internal combustion engine mechanical performance. The research shows that the input layer independent variable with the strongest correlation with the engine performance is the ignition advance angle, and the global correlation coefficient is 0.946; the correlation coefficients between the five oil products and the engine performance are all less than 0.2, which has a very weak correlation.