遗传-神经网络算法优化飞机垂尾

通过优化垂尾复合材料蒙皮铺层体系，可以减少垂直尾翼的气动变形，进而改善飞机的方向安定性。采用遗传算法能够较好地分析这类复杂优化问题，但遗传算法的计算工作量大，计算所需时间过长。本文提出遗传一神经网络的算法，该算法首先采用神经网络预测遗传变量的值，进而缩短遗传变量的范围，然后采用遗传算法优化得到最优解。结果表明，这种方法在保证计算精度的情况下，减少了计算时间，提高了工作效率。

Optimal design of vertical tail using genetic-neural network algorithms

The direction stabilizer of aircraft can be improved by optimizing the composite laminates of vertical stabilizer.Although genetic algorithms can settle this complex optimal problem,it takes too much computational time.So,the genetic-neural network algorithms was proposed to optimize the skin laminate in this paper.First,neural network was used to predict the value of genetic variable,so the range of genetic variables can be greatly reduced.Finally, genetic algorithm is used to obtain the best solution.The result shows that by using this method computing time was reduced meanwhile computation precision was assured.