CFA6470HEV发动机燃油经济性的BP神经网络建模及优化

BP Neural Network Modeling and Optimization on Engine Fuel Economy of CFA6470HEV

对MATLAB软件中的BP神经网络及数值分析方法进行了研究,并将其应用于CFA6470HEV发动机特性的建模,在对模型的结构参数优化的基础上,获得了合适的发动机特性模型。应用数值分析方法确定了发动机的最优经济性运行区间以及与之对应的发动机节气门开启角与转矩和转速之间的关系,利用此关系,可通过对发动机节气门的调节实现对发动机最佳燃油经济性的控制。

The back-propagation （BP） neural network and numerical analysis methods in MATLAB soft- ware are studied and applied to the modeling of the CFA6470HEV engine characteristics. Based on the structural parameters optimization of the model, an appropriate model for engine performance is obtaind. By using numerical analysis methods, the engine operation range with optimal fuel economy and the corresponding relationship among throttle opening, torque and rotation speed of the engine are determined. With the relationship, the engine control for optimal fuel economy can be achieved by adjusting the throttle opening of the engine.