空天飞行器上升轨迹优化  

Ascent Trajectory Optimization of Aerospace Vehicle

作　　者：张凯[1] 石文[1] 郑安波[1] 付义伟[1] 

机构地区：[1]中国航天空气动力技术研究院,北京100074

出　　处：《航空计算技术》2012年第6期62-64,69,共4页Aeronautical Computing Technique

摘　　要：飞行器轨迹优化是一类最优控制问题,传统的优化方法存在对初值敏感的缺陷。采用一种组合优化方法,对空天飞行器的上升轨迹进行优化。首先用遗传算法对直接打靶法离散的非线性规划问题进行全局寻优,得到初步优化结果,然后以该结果作为初值,用序列二次规划法对伪谱法离散的非线性规划问题进行局部寻优,得到最终的优化结果。结果表明,组合优化方法既解决了初值敏感问题,又有较高的优化精度,可以有效求解轨迹优化问题。The trajectory optimization problem is a kind of optimal control problem. Conventional direct and indirect methods are sensitive to the initial values. In order to optimize the ascent trajectory of aerospace vehicle, a combined optimization method was used. First, direct shooting method was used to convert the trajectory optimization problem into a nonlinear programming （NLP） problem, and then the NLP problem was solved by genetic algorithm to get initial guesses. Second, pseudo spectral method was used to transform the trajectory optimization problem into a NLP problem, and then sequential quadratic programming was used to solve the NLP problem with initial values from genetic algorithm. It is shown that the combined method is effective for the solution of trajectory optimization.

关 键 词：空天飞行器 轨迹优化 遗传算法 伪谱法 

分 类 号：V411[航空宇航科学与技术—航空宇航推进理论与工程]