【作者】 张侃；

【导师】 刘燕斌；

【作者基本信息】 南京航空航天大学 ， 导航、制导与控制， 2010， 硕士

【摘要】 空天飞行器结合了航空航天技术,具有飞行速度快、重复使用率高和功能齐全的优点,使其不仅具有多种民用用途而且还具有较高的军事价值,它的推出势必会引起新一轮的军事革命。基于此,本文重点研究了空天飞行器的轨迹设计与优化问题。空天飞行器轨迹优化问题本质上是一类最优控制问题,针对该问题,本文提出模型变换和最优化计算的分步求解思路,分别讨论每一步中不同方法的优缺点,并给出求解步骤。建立了空天飞行器的三自由度运动模型,并精确分析影响最优轨迹精度的环境模型、大气密度模型、声速模型和地球引力模型等。研究了空基投放型空天飞行器的规避飞行轨迹优化问题,根据任务把飞行过程分为上升段和巡航段,提出了分段建模,分段优化的思想。针对传统的间接法最优性条件推导繁琐,最优轨迹对初值比较敏感的缺陷,首先利用直接配置法将无限维的动态优化问题转换成有限维的静态优化问题,即非线性规划问题,然后考虑到非线性规划问题约束条件复杂,可行域较窄,提出了逐步细分的优化策略,并利用序列二次规划算法对其求解,最后进行了仿真验证和误差分析。研究了空天飞行器的滑翔轨迹优化问题,提出了基于Legendre伪谱法的模型变换方法,设计了遗传算法和序列二次规划相结合的串行组合优化策略,通过计算轨迹安全走廊得到了初始参考轨迹,加快了收敛速度,最后基于最远航程性能指标的仿真实例说明了Legendre伪谱法同样可以很好地处理轨迹优化问题,设计的串行组合优化策略计算效率高,切实有效。更多还原

【Abstract】 As the combination of aviation technology and space technology, aerospace vehicle has the advantages of fast flight speed, high reuse rate, strong mobility and complete functions, such that it not only has many civilian uses, but also has high military, and may cause a new round of military revolution. Therefore, this paper focuses on the trajectory design and optimization of aerospace vehicle.In essence, trajectory optimization is regarded as an optimal control problem. For this problem, the solving ideas of model transformation and optimization calculation are proposed in this paper. And the advantages and disadvantages of different methods in each step are discussed, then further the solution procedure is given.The 3DOF motion model of aerospace vehicle is established, and they are analyzed accurately to environment model, atmospheric model, sound velocity model and gravity model affecting the the optimal trajectory precision.In this paper, the dodge-flying trajectory optimization problem for an air-based aerospace vehicle is studied. According to mission requirements, the whole flight is divided into two segments: ascent phase and cruise phase, and the ideas of modeling and optimization by step are proposed. For the traditional indirect methods, whose optimal conditions derive cumbersome and whose optimal trajectory are sensitive to the initial values, firstly the direct collocation methods are used totransfer the infinite dimensional dynamic optimal problem into finite dimensional static optimal problem, namely nonlinear programming problem (NLP). Secendly considering that the constraints of the NLP problem are complex and the feasible region is narrow, a sequential-division optimization strategy is proposed to solve the optimal problem using sequential quadratic programming (SQP), finally then the simulation and error analysis are done.The glide trajectory optimization problems for aerospace vehicle are studied, and the new model transforming method based on the Legendre pseudospectral method is developed, then further a serial hybrid optimization strategy combined genetic algorithm with SQP is designed. By calculating the safe corridors, the initial reference trajectory is obtained, which can speed up the convergence. Finally, the simulation results based on the performance of maximum range demonstrate that the Legendre pseudo-spectral method can also deal with trajectory optimization problems well, and the serial hybrid optimization strategy is efficient and feasible.更多还原