An Improved Levy Flight Algorithm For Aircraft Airfoil Optimization Design

The wing is the main component of the aircraft’s lift.The optimized design of the airfoil is one of the key issues to improve the aerodynamic performance of the aircraft.The design of advanced aircraft requires the development of high-performance airfoils.In the design of airfoil based on computational fluid dynamics,efficient optimization of design methods is particularly important.In this paper,aiming at the optimization design of aircraft airfoil,the improvement of Levi’s flight optimization algorithm and Gaussian process agent model are studied.First,Levi’s flight is a global search algorithm,but the local search ability is poor.Aiming at this problem,this paper combines Levi’s flight with two efficient local search algorithms,and proposes a teacher-student exchange algorithm based on improved Levi’s flight and a wolf group algorithm based on improved Levi’s flight.At the same time,in order to overcome the shortcomings of the original Levi’s low flight efficiency and poor precision,Gaussian kernel function is introduced to adaptively adjust the search step size of Levi’s flight to increase its search efficiency.The quaternary method is introduced to increase its search accuracy.The standard test function verifies that the improved algorithm effectively balances the local search and global search ability of the algorithm,and increases the convergence speed while ensuring the global search ability of the algorithm.At the same time,the algorithm is superior to the original algorithm in terms of optimization efficiency and optimization accuracy.In the genetic algorithm.Secondly,the machine model is used to construct the proxy model of airfoil geometric parameters and aerodynamic performance.Aiming at the shortcomings of poor prediction accuracy of static proxy model,this paper constructs dynamic proxy model by using EI plus point criterion to improve the accuracy of proxy optimization,and proposes an airfoil proxy optimization method based on dynamic Gaussian process and efficient global optimization algorithm.Finally,taking the drag reduction optimization design of supercritical airfoil RAE2822 as an example,the direct optimization method and the agent optimization method are used to optimize the drag reduction.The results show that the efficiency of agent optimization is greatly improved compared with direct optimization.It is proved that the proposed airfoil proxy optimization method based on dynamic Gaussian process and efficient global optimization algorithm is feasible.