Modeling Fuel Injection Control Maps Using Fuzzy Logic and Neural Networks

Abstract

Determining the correct ignition point of the air-fuel mixture is critical in order to achieve maximum output torque and to reduce exhaust emissions. In some fuel injection control systems the amount of air cannot be detected, thus, look-up tables are utilized, which contain the amount of air for given engine speed and inlet manifold pressure. In this paper, we model the look-up table using fuzzy logic. A neural network approach is used to identify the inputs of the fuzzy model.