北京航空航天大学学报 ›› 2020, Vol. 46 ›› Issue (3): 579-587.doi: 10.13700/j.bh.1001-5965.2019.0262

• 论文 • 上一篇    下一篇

一种高超声速滑翔再入在线轨迹规划算法

李俊, 江振宇   

  1. 国防科技大学 空天科学学院, 长沙 410073
  • 收稿日期:2019-05-28 发布日期:2020-03-28
  • 通讯作者: 江振宇 E-mail:jiang_nudt@sina.com
  • 作者简介:李俊,男,硕士研究生。主要研究方向:飞行器动力学与制导;江振宇,男,博士,教授,硕士生导师。主要研究方向:飞行器动力学与制导。

Online trajectory planning algorithm for hypersonic glide re-entry problem

LI Jun, JIANG Zhenyu   

  1. College of Aerospace Science and Engineering, National University of Defense Technology, Changsha 410073, China
  • Received:2019-05-28 Published:2020-03-28

摘要: 为了提高滑翔再入飞行器响应动态任务的能力并提高其制导系统的鲁棒性,建立了高超声速滑翔再入轨迹规划问题的一种在线求解算法。该算法以一列凸子问题来逼近原问题。通过选择以弧长为自变量并引入对数速度代替速度作为状态,使得动力学方程的非线性大为减弱,同时使动压和热流约束完全成为线性约束。通过使用类似于混合整数规划中割平面的思想处理禁飞区约束,尽可能避免了不必要的计算。通过直接以气动系数和大气密度的乘积作为控制量,构造伪线性控制模型,进一步减弱非线性。非凸约束被适当松弛,以保证子问题的可行性。为了避免过度松弛,利用离线给定的高度和速度上下边界估计对应的参数,以加速收敛。以X-33再入任务为例验证了所提算法的有效性。该算法可以以简单的常值函数为初值并在数次迭代后收敛。

关键词: 再入轨迹规划, 对数速度, 凸优化, 最优控制, 割平面法

Abstract: In order to improve the capability of the glide reentry vehicle to respond to dynamic missions and improve the robustness of its guidance system, an online solving method for hypersonic glide re-entry trajectory planning problem is established. The method approximates the original problem by a series of convex subproblems. By choosing arc length as the independent variable and introducing logarithmic velocity as the state instead of velocity, the nonlinearity of the dynamic equation is greatly reduced, and the dynamic pressure and heat flow constraints completely become linear constraints. No-fly zone constraints are treated by using a cutting plane method like mixed integer programming to avoid unnecessary calculation as much as possible. By taking the product of aerodynamic coefficient and atmospheric density as the control quantity directly, the pseudo-linear control model is constructed to further weaken the nonlinearity. The non-convex constraint is relaxed appropriately to ensure the feasibility of the subproblem. In order to avoid excessive relaxation, the upper and lower boundaries of the given off-line height and speed are used to estimate the corresponding parameters in order to accelerate convergence. The X-33 re-entry task is taken as an example to verify the effectiveness of the method. The method can take simple constant function as initial value and converge after several iterations.

Key words: re-entry trajectory planning, logarithmic velocity, convex optimization, optimum control, cutting plane method

中图分类号: 


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发