Optimizing unmanned aerial vehicle aerodynamics: a two-stage approach using genetic algorithms and adjoint methods

Abstract

In this article, we discuss the design of a flying wing using a default model and the application of the Taguchi test design method, genetic evolutionary optimization algorithm, and adjoint derivative optimization method, along with fluid dynamics calculations. Simulations were conducted using compressible Reynolds-averaged equations based on the Large Eddy Simulation (LES) turbulence model and the k-ω Shear Stress Transport (SST) model at a speed of 50 m/s and a Reynolds number of 1.6 million. The first stage of optimization was performed on the wing sections using the adjoint method. Subsequently, the second and third stages were optimized through a response surface method and genetic algorithm, involving 11 variables across these stages with a total of 50 tests conducted. The test case is a lambda-shaped flying wing named Saccon, which features wings with a 53-degree sweep angle. In this research, in order to optimize the lift coefficient and the ratio of lift to drag coefficients, a target function has been defined for the lift-to-drag ratio. By selecting this objective, both the lift and the lift-to-drag ratio can be increased. The results indicate that the impact of the variables varies for each section of the wing. Ultimately, the newly designed body achieved a 10% increase in aerodynamic efficiency compared to the default model through the use of optimization algorithms. Additionally, the drag coefficient increased by 140%, successfully completing the design process with multi-stage optimization techniques.