重型运载火箭轨迹/总体参数一体化优化方法

An Integrated Optimization Method of Trajectory/System Parameters for Heavy Launch Vehicles

研究了重型运载火箭轨迹/总体参数一体化优化问题。建立了一体化参数优化模型,设计了一种双环迭代优化算法,外环采用遗传算法进行参数优化,内环通过带自适应松弛因子的改进牛顿迭代法进行参数迭代。针对牛顿迭代法对模型本身及迭代初值敏感的问题,提出了一种模型变量筛选方法和迭代初值生成方法。基于Morris参数全局灵敏度分析法筛选出对终端等式约束影响最大的参数作为牛顿迭代变量,以历史迭代收敛值作为当前迭代初值。并通过迭代误差监控,保证了算法的鲁棒性。以三级液体重型运载火箭为对象,进行了一体化优化仿真。仿真结果显示起飞总质量减少了4.09%,表明设计的双环迭代优化算法能在复杂约束下完成重型运载火箭轨迹/总体参数一体化优化。

The integrated optimization of trajectory / system parameters for heavy launch vehicles was investigated. A model of integrated optimization was set up. An iterative optimization algorithm with double loops was designed. The genetic algorithm was used to optimize parameters in outer loop,and the improved Newton iteration method with an adaptive relaxation factor was used to iterate parameters in inner loop. Aiming at the susceptibility of Newton iteration method to the model and initial values,a method of model variable screening and initial iteration values generation were proposed. Based on the Morris method of global sensitivity analysis,the parameters with the biggest influence on the terminal equality constraints were taken as the Newton iteration variables. And the historic convergent iteration values were set as the current initial iteration values. And the robustness of the algorithm was ensured by monitoring of the iterative error. A liquid heavy launch vehicle with 3 levels was set as an example for integrated optimization simulation. The results showed that initial mass decreases by 4.09% and demonstrate that the iterative optimization algorithm with double loops can complete the integrated optimization of trajectory / system parameters for heavy launch vehicles with complex constraints.