For a returnable hypersonic cruise vehicle, the whole trajectory design is strongly coupled to the cruise state. To deal with the coupling relationship, a bilevel trajectory optimization problem is proposed in this paper. The upper level is for single-objective optimization of the whole trajectory, whereas the lower level is to conduct multiobjective optimization of the cruise flight. Based on the bilevel directed search domain (BDSD) algorithm, a bilevel trajectory optimization approach is proposed. Firstly, the dynamics model is established according to the coupling among the propulsion, aerodynamics, and trajectory. Then, the coupling relationships among different flight phases and those between the cruise trajectory and operate window of the scramjet are converted to optimization constraints. Thus, the highly coupled bilevel trajectory problem is transformed into a decoupled bilevel optimization problem, enabling the BDSD algorithm to solve the problem. Finally, the availability of the proposed approach is verified by the numerical results. Compared with another efficient method, the approach presented in this paper can drastically reduce the optimization time from tens or even hundreds of days to only about 8 days. Because of the high efficiency of the approach, the iterative trajectory design at the conceptual design stage is possible.

对于可返回的高超音速巡航飞行器，整个轨迹设计与巡航状态密切相关。为了解决耦合关系，提出了一种双层轨迹优化问题。上层是对整个轨迹的单目标优化，而下层是对巡航飞行进行多目标优化。基于双层有向搜索域（BDSD）算法，提出了一种双层轨迹优化方法。首先，根据推进力，空气动力学和弹道之间的耦合建立动力学模型。然后，将不同飞行阶段之间以及超燃冲压发动机的巡航轨迹和操作窗口之间的耦合关系转换为优化约束。从而，将高度耦合的双层轨迹问题转换为解耦的双层优化问题，从而使BDSD算法能够解决该问题。最后，数值结果验证了该方法的有效性。与另一种有效方法相比，本文提出的方法可以将优化时间从数十天甚至数百天大幅减少到仅8天左右。由于该方法的高效率，因此可以在概念设计阶段进行迭代轨迹设计。本文介绍的方法可以将优化时间从数十天甚至数百天大幅减少到仅约8天。由于该方法的高效率，因此可以在概念设计阶段进行迭代轨迹设计。本文介绍的方法可以将优化时间从数十天甚至数百天大幅减少到仅约8天。由于该方法的高效率，因此可以在概念设计阶段进行迭代轨迹设计。