自動車エンジン排気系形状の多目的最適化(流体工学, 流体機械)

抄録

To develop a better car engine, many design specifications are required because they would affect intake/exhaust system performance. Such specifications have to be found out by trial and error through many experiments and analysises. Therefore the automated multiobjective design optimization is desirable to reduce technical, schedule, and cost risks. Exhaust air has to be led from piston chambers to the exhaust pipe smoothly to increase the engine's power. It also needs to keep high temperature in the exhaust pipe because the catalyst located at the end of exhaust pipe will absorb more pollutant in higher temperature conditions. The design optimization has multiple objectives. This paper presents the development of the optimization code using Multi Objective Genetic Algorithm (MOGA).