The Research on Air Fuel Ratio Predictive Model of Gasoline Engine during Transient Condition

In order to overcome air fuel ratio transmission delay influence on air fuel ratio control accuracy of gasoline engine during transient conditions, a new multi-step predictive model of air fuel ratio based on BP neural network was provided in this paper. Input vectors of neural network multi-step predictive model were determined by the mathematic model of air fuel ratio, and the derivation of air fuel ratio reflecting the air fuel ratio tendency was included within input vectors of neural network to improve the prediction accuracy of air fuel ratio model. The simulation was accomplished using the experiment data of HL495 gasoline engine during transient conditions, and weight values of back propagation neural network were adjusted by gradient algorithm and adaptive method, and the results show the maximal error of air fuel ratio predictive model is less than 3% and the average error of that model is less than 2%, and the air fuel ratio predictive model can accurately approximate air fuel ratio process of gasoline engine during transient condition.