摘要
液体火箭发动机推进弹道式导弹总体设计参数的全局最优化问题是亟待解决的计算问题。遗传算法具有全局搜索能力强、鲁棒性强、适于并行处理的特点,而Powell算法具有很好的求解局部最优解的能力。将两种方法进行有效改进后使之相结合,设计出并行全局最优化混合遗传算法。并以此为基础,建立了液体火箭发动机推进弹道式导弹总体优化设计模型。以液体火箭发动机推进弹道式导弹的起飞质量最小为目标,对液体推进剂弹道式导弹设计参数进行了优化设计。数值优化结果表明:该混合算法提高了搜索全局最优解的速度,优化精度高,且避免了初值敏感、病态梯度和局部收敛等问题,能够搜索到全局最优设计参数。
The matters of global optimization of configuration design for liquid rocket engine propelled ballistic missiles were in dire need of computation. Genetic algorithms have the characters such as strong ability in find global optimal solution, strong robustness, and good parallel efficiency, while Powell algorithm had very strong ability to find the locally optimal solution. In this paper, two methods after some efficient improvement were combined, and a parallel global optimization hybrid genetic algorithm was designed. A numerical optimization model for liquid rocket engine propelled ballistic missiles is developed based on the hybrid genetic algorithm. Take-off weight of liquid rocket engine propelled missiles was taken as the objective function and the design parameters for liquid rocket engine propelled missiles as optimization design variables. The numerical optimization results show that this hybrid algorithm could efficiently get the globally optimal solution and take on high optimization precision. Moreover, it avoids many shortcomings, such as initial value sensitivity, ill-conditioned gradient and local convergence, and so on. It can obtain global optimum design parameters.
出处
《飞行力学》
CSCD
北大核心
2011年第3期44-47,51,共5页
Flight Dynamics
