混合遗传算法及其在叶片自动优化设计中的应用

 在遗传算法中 ,采用单纯形法寻优取代变异运算构成混合遗传算法 ,以提高遗传算法局部搜索能力。算例表明混合遗传算法可有效提高搜索效率和对最优解的逼近程度。采用三次多项式和多圆弧方法生成叶型中弧线 ,三次多项式分布叶型厚度 ,对叶型进行参数化。将N S方程正问题流场数值计算与混合遗传算法相结合 ,构成叶轮机械叶型自动优化设计。尝试由气流转角、总压损失和叶型型面面积构成目标函数 ,对压气机叶型进行自动优化设计。

COMPOSITE GENETIC ALGORITHM AND ITS APPLICATION TO AUTOMATIC OPTIMIZATION OF TURBOMACHINERY BLADE

To improve the locally searching ability of a genetic algorithm, the mutation operation is replaced by the process of searching for an optimum using the simplex method. The examples show that using the composite genetic algorithm can effectively improve the search efficiency and approach degree to the optimal objective. For constructing the blade mean camber lines, a third order polynomial and a multi arc curve is used. The blade thickness is distributed also by another third order polynomial. The composite genetic algorithm is combined with 2DNS flow field calculation, for an automatic optimization design of the turbomachinery blade. It is attempted that constructing an objective function by the flow turning angle, total pressure loss and blade area; then the composite genetic algorithm is used to design a compressor blade.