基于并联协作混合遗传算法的高超声速巡航飞行器一体化优化设计研究

Parallel collaborative hybrid genetic algorithm applied to integrated design optimization of hypersonic cruise vehicle

遗传算法能以很高的概率找到全局最优解 ,不需要目标函数和约束条件的梯度信息 ,适合于工程优化设计。为增强遗传算法的局部搜索能力 ,提出了一种新的融合模式搜索方法和 Powell方法的并联协作混合遗传算法。研究超燃冲压发动机为动力的高超声速巡航飞行器的一体化优化设计 ,系统分析了机身和超燃冲压发动机设计参数对飞行器总体性能的影响。通过建立高超声速巡航飞行器质量模型、气动力估算模型、气动热估算模型、超燃冲压发动机性能分析模型、控制模型和弹道分析模型 ,在满足参考任务要求的前题下 ,以飞行器起飞质量为目标函数 ,将并联协作混合遗传算法应用于最优总体方案和超燃冲压发动机方案一体化设计 ,完成了 6设计变量的全局优化计算。算例表明 ,本文建立的一体化设计模型基本正确 ,并联协作混合遗传算法是用于高超声速巡航飞行器一体化优化设计的较好的优化算法。

Genetic algorithm (GA) did not require gradient information of objective function or constraints.As exploring most of the design space stochastically,GA was very likely to find the global optimum and suited for engineering optimal designs.A new kind of hybrid GA combining patter search and Powell methods in parallel collaborative mode had been proposed to improve the local search ability of standard GA.In this paper,an integrated optimal design process of hypersonic cruise vehicle propelled by scramjet was presented,and the influences of design parameters of airframe and scramjet on vehicle performance were analyzed.By developing mass properties,aerodynamic forces,aeroheating,scramjet,control and trajectory performance models,considering of the reference mission design requirements,an optimization process of 6 design parameters basing on the newly developed parallel collaborative hybrid GA was performed to minimize gross lift-off weight (GLOW) of hypersonic cruise vehicle.Numerical results showed that the integral design model was primarily right and hybrid genetic algorithm was an ideal algorithm for the integral design optimization of hypersonic cruise vehicle.