基于多岛遗传-模拟退火算法的气动外形优化

针对超声速翼型滑翔机的气动外形优化问题,提出了一种多岛遗传算法与模拟退火算法相结合的混合优化算法。首先,通过多岛遗传算法产生一个随机的初始种群,找到全局最优点附近区域的一个次优解;然后,将此次优解作为模拟退火算法的初值启动退火进程,缩小设计空间的范围,找到全局最优解。优化结果表明,所提混合算法可以有效地解决飞行器外形优化问题;在数量巨大的设计空间中,滑翔机的最优气动外形能够以较低的计算资源代价快速得到;双弧形翼型的气动特性较六边形翼型更有优势,叉形尾翼的静稳定性和控制效率高于十字形;优化后的外形极大地增加了滑翔机的滑翔距离。

Aerodynamic shape optimization based on the hybrid MIGA-SA method

For the aerodynamics shape optimization of the supersonic glider,this paper presents a hybrid optimization algorithm combining the multi-island genetic algorithm with the simulated annealing methods.Firstly,a random population is generated and a suboptimal solution is found near the global optimal solution area through MIGA.Then,it takes the suboptimal solution as SA's initial value and starts annealing process of SA,narrow the range of the design space,and search the area to find the global optimal solution.The optimization results show that the improved hybrid method can effectively solve the flight vehicle optimal design issues.The glider optimal aerodynamic geometry can be quickly obtained by using of the hybrid method under the conditions of huge design space and low calculating resource requirements.The aerodynamic characteristics for the double curved airfoil are always superior to the hexagonal airfoil,the static stability and the control efficiency of the forked configuration tail are higher than the cross configuration.The final optimal geometry can greatly extend the flight distance for the glider.