A Multi-disciplinary design and optimization platform for transport airplane
with environmental considerations is developed. Instead of employing the usual
optimization approach that minimizes the averaged direct operational cost in a set of
stage lengths, an enhanced multi-disciplinary methodology involving a considerable
detailed airplane model is presented here. In the past, the suitability of the product for a
realistic airline network has not been entirely considered, and even disconnected from
airline real needs. Increased computer power, sophisticated surrogate models, and new
optimization techniques make possible to optimize the complete system with aircraft
inserted in an air transport network including complex mission analyzes. In this
context, a sophisticated airplane design framework with multi-disciplinary approach
encompassing several aeronautical disciplines including noise and emissions was
elaborated for the design of optimal airplane configurations that are suited for a desired
airline network. Artificial neural network is used as a surrogate model to calculate the
airplane aerodynamics with a high-fidelity approach, to perform a realistic mission
analysis enabling operation aspects to properly be considered in the optimization task.
The simultaneous minimization of total network direct operational costs and
maximization of Total Network Yield is performed to optimize fifteen aircraft design
variables, encompassing airframe and engine definition. The MOGA-II genetic
algorithm, available in the modeFrontier® Integrated Optimization Environment is
employed in the present design simulations
Keywords: Aerial network, Aircraft conceptual design, Applied aerodynamics,
Artificial neural network, Evolutionary computation, Multi-disciplinary design
and optimization, Propulsion, Stability and control.
