垂直起降可重复使用火箭返回着陆过程实时轨迹优化

Reusable launch vehicle is crucial for sustainable development of the aerospace industry. Recovery of the rocket is a significant prerequisite for its reuse. Recently VTVL (Vertical Takeoff and Vertical Landing) has become the most promising recovery technique. However, the return and landing process of VTVL rocket presents challenges to guidance and control, including severe coupling between aerodynamic force and thrust, rapidly changing flight environment, considerable model uncertainty, harsh landing conditions, etc. This research focuses on real-time trajectory optimization for descent and landing of VTVL reusable rockets to realize safe, stable and economic flight along with high-precision landing. The process of interest is characterized by multiple stages and nonlinearities, with constraints on flight angular rate, angle of attack, sideslip angle, overload, fuel, engine tilt angle, ignition conditions, as well as on the terminal velocity, position and attitude. Based on the above knowledge, topics to be investigated include mechanism modeling and optimization-oriented formulation for returning and landing of VTVL reusable rockets, multi-stage nonlinear dynamic optimization with complex constraints, and moving-horizon online trajectory optimization. The models and algorithms will be developed with expected features of comprehensive description, simultaneous solution and real-time computing capability. This work will be verified on a certain type of launch vehicle in China.

可重复使用运载火箭的研发和商业化将极大推动航天工业的快速、可持续发展。目前垂直起降(Vertical Takeoff and Vertical Landing, VTVL)已成为火箭回收最具前景的方式之一。火箭返回着陆过程存在气动力和推力复合控制耦合严重、飞行环境大幅快变、模型参数不确定度大、着陆条件苛刻等制导控制难题。为保证返回过程的安全、平稳、经济飞行，并确保高精度着陆，本项目基于返回着陆过程的多阶段、动态非线性特性，综合考虑火箭返回着陆过程的飞行角速率、攻角、侧滑角、过载、燃料、发动机摆角、发动机点火条件等过程约束，以及着陆点速度、位置、姿态等终端约束，研究垂直起降可重复使用火箭返回着陆过程的机理模型构建与优化命题构造、多阶段复杂约束下的非线性系统动态优化、滚动时域轨迹优化等科学问题。将形成整体表达、联立求解、实时计算为特征的优化模型和算法，并依托某型运载火箭对象模型进行仿真验证。

本项目着重对可重复使用运载火箭技术研发中存在的气动力和推力复合控制耦合严重、飞行环境大幅变化、模型参数不确定度大、着陆条件苛刻等问题展开深入研究。基于返回着陆过程的多阶段、动态非线性特性，综合考虑火箭返回着陆过程的飞行角速率、攻角、侧滑角、过载、燃料、发动机摆角、发动机点火条件等过程约束，以及着陆点速度、位置、姿态等终端约束，构建了可重复使用火箭返回着陆过程的机理模型与优化命题。进一步，通过滚动时域优化、序列凸优化、在线实时优化计算等技术，针对重复使用运载火箭在不同场景下的优化命题进行了快速优化计算，同时保障了求解优化命题的收敛性与精确性。大量的仿真实验结果验证了提出算法的可行性、鲁棒性与高效性。