An Example Of Genetic Algorithms Applied To Multidisciplinary Design Optimization

José Alexandre Tavares Guerreiro Fregnani, Bento S. de Mattos (Technological Institute of Aeronautics)

CHALLENGE - This study investigates the use of Genetic Algorithms (GA) algorithms as main optimization tool in a Multidisciplinary Design and Optimization (MDO) framework considering three disciplines (aerodynamics, propulsion and aircraft performance) for a given set of aircraft parameters (airframe and engines geometries), inserted in each operational scenario. The mission analysis is evaluated into a 3D payload-range-DOC envelope generated through this framework.

SOLUTION - A MDO calculation framework is designed for a given aircraft configuration (engines and airframe) and operational conditions with the main objective is to determine the optimum mission flight path (optimum flight speeds and altitudes), minimizing the Direct Operational Cost (DOC, in US$/nm) and maximizing Zero Fuel Weight (ZFW, in kg) for a mission range interval. A generic regional transport jet, designed to transport 78 passengers, is used in the study. The calculation framework is composed by three calculation modules (aerodynamics, propulsion and mission performance) integrated into a Genetic Algorithm optimizer. These modules are designed on MATLAB code and the integration with the GA optimizer is done via modeFRONTIER software.

BENEFITS - The calculation framework using MOGA-II algorithm was executed with 900 individuals in 30 designs (generations) in a total processing time of 182min. Feasible designs (not restricted) represented 95.3% of the total individuals generated.