可重复使用助推器翻转垂降回收研究

以火箭助推器回收为背景，提出一种全新的翻转回收方式，避免其初始姿态大范围调整，有效节约燃料和简化设计。首先，基于动力学建模，研究基于神经网络的着陆落点预测方法，以及满足落点精度的减速发动机最优点火策略，以最优的状态进入返回飞行过程。然后，采用PSO初始轨迹生成和Gauss伪谱法相结合的轨迹优化算法，研究返回着陆段多约束轨迹优化问题；考虑返回过程中较大的外部扰动，研究一种基于不确定性的鲁棒轨迹规划方法；同时为了提高轨迹规划的快速性，实现模型或环境突变情况下的在线轨迹规划，研究一种基于模型预测和凸优化算法的快速轨迹规划方法。第三，为保证返回降落制导控制精度，考虑助推器末端垂直降落时多约束和强扰动，研究多约束下的最优鲁棒制导律和随机鲁棒优化姿态控制律，同时还采用制导控制一体化技术提高助推器垂降落点、速度和姿态的控制精度。最后，设计仿真软件平台和无人机翻转垂降演示平台，对上述方法进行仿真和演示。

Based on the background of rocket booster, a new method of reverse recycling is put forward to avoid the large range adjustment of the initial attitude, save fuel and simplify the design. First of all, based on dynamics modeling, the landing point prediction method based on neural network, and the reduction of placement accuracy of fire engine the most advantage strategy, to the best state into the return flight. Then, the trajectory optimization algorithm combined with PSO initial trajectory generation and Gauss pseudo-spectral method is adopted to study the multi-constraint trajectory optimization of the landing segment. Considering the large external disturbance in the return process, a robust trajectory planning method based on uncertainty is studied. Meanwhile, in order to improve the rapidity of trajectory planning, to realize the online trajectory planning of model or environment mutation, a rapid trajectory planning method based on model prediction and convex optimization algorithm is studied. Third, in order to ensure the return to landing guidance control precision, considering the booster end constraint and the strong disturbance, vertical landing is studied under the constraint of the optimal robust guidance law and the stochastic optimal attitude control law. The control precision of the vertical landing point, speed and attitude of the booster is improved by using the integrated guidance control technology. Finally, the design simulation software platform and uav inverted vertical drop demonstration platform are used to simulate and demonstrate the above methods.