Applying genetic algorithms to decision making of injection strategies in a GDI engine

The effects of single or double injection strategies on the performance of a 4-cylinder, 4-valve, 4-stroke GDI engine are numerically analyzed. A properly developed 3D engine model is employed. The runs of the 3D model are driven by the MOGA-II algorithm to search for the injection strategy and spark advance that are shown optimal for the combustion development. After the optimal injection pressure for a single injection event is identified, the gasoline supply is considered as subdivided into two successive parts, each delivering the same mass of gasoline. The optimization algorithm is used to explore the actual advantages of modulating the mixture formation for low consumption and reduced environmental impact. The performed computations also allow highlighting interesting aspects, as the adverse phasing of the engine occurring under certain circumstances. ©Civil-Comp Press, 2013.