数学优化
非线性规划
线性化
轨迹优化
计算机科学
弹道
凸优化
最优化问题
线性规划
正多边形
非线性系统
数学
最优控制
物理
天文
量子力学
几何学
作者
Rouhe Zhang,Naigang Cui
标识
DOI:10.1109/taes.2023.3319590
摘要
Trajectory optimization of a hypersonic glide vehicle (HGV) has a highly nonlinear model and complex constraints. Therefore, it is difficult to deal with polygonal no-fly zone (PNFZ) constraints that are common in actual flight environment of the HGV. To solve this problem, this article proposes a general PNFZ avoidance (GPNA) algorithm to solve the HGV trajectory optimization problem with general PNFZ (convex and concave PNFZs) constraints. To the author's knowledge, this problem has never been solved before. First, the trajectory optimization problem with the convex PNFZ constraint is considered. Inspired by sequential convex programming (SCP), direct linearization and relaxation techniques are used to transform this problem into a sequence of mixed-integer linear programming problems, which can be solved efficiently by state-of-the-art software. Different from the convexification process of SCP, this article also linearizes the nonlinear convex functions. At the same time, PNFZs of actual mission are usually concave, so concave PNFZ is also considered. The convex decomposition technique is used to decompose it into multiple convex PNFZs so that the GPNA algorithm can solve more practical problems. In addition, we also consider inter-sample obstacle avoidance constraints. Simulations in various scenarios verify the effectiveness of the GPNA algorithm. The GPNA algorithm has strong adaptability and high computational efficiency, and can generate trajectories avoiding the general PNFZ within 10 s, which has a strong application value.
科研通智能强力驱动
Strongly Powered by AbleSci AI