Contamination in the world, especially those from automobile emissions is increasing rapidly day to day. Catalysts are designed to optimize the fuel and air ratio but the air fuel ratio is very sensitive to the engine's operating mode. This paper presents a solution based on machine learning technique to control the input parameters of the fuel and air ratio in electronic fuel injection engine to reduce harmful emissions that pollute the environment. Automobile engine models and neural-based controllers rely on neural networks to minimize emissions generated in fuel injection engines are designed and tested. Finally, the paper offers neural controller has been to optimize performance, limits the necessary control action.

No abstract