遗传与枚举混合算法对航空发动机复合材料主轴的优化设计

针对复合材料主轴优化设计中离散与连续变量共存的问题,提出了一种遗传算法与枚举法相结合的混合算法.以某航空发动机低压涡轮轴为原型,在满足静强度指标、临界转速及外廓尺寸要求前提下,利用该混合算法对复合材料主轴的铺层角度与厚度进行优化设计,以减轻主轴质量.结果表明:对于1~10层铺层,相比于全排列算法与单一遗传算法,混合算法能再减轻质量5.54%与3.87%;且当铺层数大于6时,混合算法计算效率能最多提高3.1倍.最后分析给出复合材料铺层厚度受强度的限制,不宜太薄的建议.

Optimizing design of composite spindle of aero-engine with hybrid algorithm consisting of genetic algorithm and enumeration algorithm

For the case of coexistence of discrete and continuous variables in optimizing design of composite spindle, a hybrid algorithm consisting of genetic algorithm and enumeration algorithm was presented. From the prototype of an aero-engine low pressure turbine shaft, this hybrid algorithm was used to optimize ply angles and thickness of composite spindle for losing the weight of spindle under the premise of achieving the targets of static strength, critical speed and overall dimension. The results show that the hybrid algorithm can realize 5.54% and 3.87% extra weight loss for from 1 to 10 plies compared with full permutation algorithm and single genetic algorithm; and when plies numbers are greater than 6, the optimizing efficiency can be improved up to 3.1 times with using hybrid algorithm. Finally, the suggestion is given that the composite layer thickness is limited by strength and should not be too thin.