遗传算法在风洞实验优化中的应用研究

为了提高风洞实验效率,降低实验成本,缩短实验周期,笔者探讨了将遗传算法引入到风洞优化实验中,实现了基于遗传算法的多段翼型实验规划.通过遗传算法对多段翼型的迎角及各段的偏转角度、重叠量和缝道宽度进行编码,由实验提供适应度值.对两段翼型的研究表明应用遗传算法规划风洞实验能够减少实验次数约40%.种群数为染色体长度的2倍时,算法能较好的搜索到最优值.初始群体值对算法的收敛性及计算效果基本无影响.此外,也模拟计算了4段翼型风洞实验,提高实验效率大约为87%～93%,可见遗传算法仍然有效且在大规模风洞实验中更有应用价值.

The application research of genetic algorithm in wind tunnel experiment optimization

To improve wind tunnel experiment efficiency, reduce experiment cost and cut experiment period, the combination of genetic algorithm with wind tunnel experiment is discussed. The attack angle, flap deflection, overlap and gap of multi-element airfoil are coded with genetic algorithm, and the fitness is provided by experiment. The research based on a 2-element airfoil shows that genetic algorithm can decrease times wind of tunnel experiment by 40 %. When the population size is twice chromosome length, algorithm can search optimum values. Initial population values almost have no influence upon the astringency of genetic algorithm and the numerical data. In addition, we also numerate the 4-element airfoil wind tunnel experiment. Algorithm can increase the experiment efficiency by 87 % - 93 %, so it may be applied to the large-scale wind tunnel experiment.