Study On Simulation Of Transient Air-fuel Ratio Control Strategy For Gasoline Engine

With the socio-economic developing rapidly, people's demands and dependences on automobile are growing, and then automobile productions and holdings have been increasing substantially. However, as a major factor on energy consumption and environment pollution, energy crisis and environmental pollution will be more severe. Especially while engine is running in transient conditions. In order to meet stringent vehicle emission regulations, study on the transient conditions of engine is imminent.According to the engine combustion equation we can know that only when the gasoline engine's Air-Fuel Ratio (AFR) infinitely close to the theoretical value, the engine can have sufficient burning and get greatest power. EFI system and three catalytic converter is a good method to minimize emissions. If we want to have the highest convert ratio of CO,HC and NOx that transformed by the three catalytic converter, we must keep the AFR in a very small range close to the theoretical value(±3. 5%). So, keeping the AFR of engine on transient conditions has important practical value.In the paper, a Mean Value Engine Model (MVEM) is created on the MATLAB/Simulink platform, which contains three sub-models: fuel flow/evaporation dynamic sub-model, manifold filling dynamics, crank shaft speed state equation. At the same time, throttle degree controller and AFR calculator sub-model have been built to control the throttle degree and calculate the AFR separately. Fuzzy controller has been established by the GUI of the Fuzzy Logic Toolbox on the MATLAB/Simulink platform. Neural Network controller and Fuzzy Neural Network controller procedures are complied with S function firstly and then established on the MATLAB/Simulink platform. After all the controllers were completed, engine's AFR control Simulation could run on transient conditions.The results show that: Comparing with fuzzy control and neural network control, fuzzy neural network control has both the advantages. When using fuzzy neural network controller, the AFR can be controlled very well and the system have shorter adjust time than that when using fuzzy controller and also has smaller Overshoot than using neural network controller. We can see that fuzzy neural network controller have excellent control performance.