To develop the control of the transient air-fuel ratio of a single-cylinder motorcycle engine of large displacement

an intake quantity estimating model based on back propagation (BP) neural network and a fuel film compensation model were built with the Matlab/Simulink

and the simulation of air-fuel ratio control was conducted with the two models. The results indicate that the intake quantity estimating model can limit the overshoot of air-fuel ratio to less than 10%

and resume the mixture to homogeneous equivalence ratio in 1.4 s

avoiding the concussion that generally takes place in conventional proportion integration differentiation control. This illustrates that the air-fuel ratio control based on transient air flow prediction works well. Its combination with the fuel film compensation can make the air-fuel ratio overshoot decline to less than 5% and the mixture resume to homogeneous equivalence ratio in 1.2 s. This shows that the introduction of fuel film compensation algorithm can significantly decrease the impact of dynamic fuel-transmission characteristic

improving the precision of air-fuel ratio control.