A Fuzzy Neural Network and Application to Air-Fuel Ratio Control under Gasoline Engine Transient Condition

In the paper, a Hendricks Mean Value Engine Model is established by using SIMULINK. At the same time, a fuzzy neural network is designed. The AFR is simulated under transient conditions. The simulation result shows that: With no controller, when throttle degree is changed intensively, the AFR errors are large, With the FNN controller, the AFR errors can be controlled to a narrow range, and the system has shorter adjust-time, smaller overshoot. So, the fuzzy neural network controller has good control performance under gasoline engine transient condition.