航空学报 > 2016, Vol. 37 Issue (10): 3149-3158   doi: 10.7527/S1000-6893.2016.0064

基于罚函数序列凸规划的多无人机轨迹规划

王祝1,2, 刘莉1,2, 龙腾1,2, 温永禄1,2   

  1. 1. 飞行器动力学与控制教育部重点实验室, 北京 100081;
    2. 北京理工大学 宇航学院, 北京 100081
  • 收稿日期:2015-11-11 修回日期:2016-01-22 出版日期:2016-10-15 发布日期:2016-03-07
  • 通讯作者: 刘莉,Tel.:010-68914534 E-mail:liuli@bit.edu.cn E-mail:liuli@bit.edu.cn
  • 作者简介:王祝 男,博士研究生。主要研究方向:无人机任务规划、无人机协同决策与控制。Tel:010-68913290 E-mail:wangzhubit@163.com;刘莉 女,博士,教授,博士生导师。主要研究方向:飞行器总体设计、飞行器结构设计、多学科设计优化。Tel:010-68914534 E-mail:liuli@bit.edu.cn
  • 基金资助:

    国家自然科学基金(11372036,51105040);航空科学基金(2015ZA72004)

Trajectory planning for multi-UAVs using penalty sequential convex pro-gramming

WANG Zhu1,2, LIU Li1,2, LONG Teng1,2, WEN Yonglu1,2   

  1. 1. Key Laboratory of Dynamics and Control of Flight Vehicle, Ministry of Education, Beijing 100081, China;
    2. School of Aerospace Engineering, Beijing Institute of Technology, Beijing 100081, China
  • Received:2015-11-11 Revised:2016-01-22 Online:2016-10-15 Published:2016-03-07
  • Supported by:

    National Natural Science Foundation of China (11372036; 51105040); Aeronautical Science Foundation of China (2015ZA72004)

摘要:

多无人机(UAVs)轨迹规划是具有非线性运动约束和非凸路径约束的最优控制问题。引入序列凸规划思想,将非凸最优控制问题近似为一系列凸优化子问题,并利用成熟的凸优化算法进行求解,以更好地权衡最优性和时效性。首先,建立了多无人机协同轨迹规划的非凸最优控制模型。然后,利用离散化和凸近似方法将其转换为凸优化问题,包括对无人机运动模型的线性化,以及对威胁规避约束和无人机碰撞约束的凸化。同时,提出了一种离散点间的威胁规避方法,保证无人机在离散轨迹点间的飞行安全。在凸优化模型的基础上,给出了基于罚函数序列凸规划求解多无人机轨迹规划的具体框架。最后,通过数值仿真验证了方法的有效性,结果表明该方法在多机轨迹规划结果的最优性和时效性都要优于伪谱法,而且优势随编队数量的增加而增大。

关键词: 无人机, 轨迹规划, 碰撞规避, 最优控制, 凸规划

Abstract:

Trajectory planning of multiple unmanned aerial vehicles (UAVs) is an optimal control problem which subjects to nonlinear motion and nonconvex path constraints. Based on the sequential convex programming approach, such nonconvex optimal control is approximated to be a series of convex optimization subproblems, which can be solved by the state-of-the-art convex optimization algorithm. A good balance between solution quality and computational tractability can then be achieved. Nonconvex optimal control model for multi-UAV trajectory planning is formulated first, and is then approximated to be a convex optimization by discretization and convexification methods. To convexify the nonconvex model, equations of motion of UAVs are linearized, and constraints of threat avoidance and inter-UAVs collision avoidance are convexified. Meanwhile, an inter-sample threat avoidance method is provided to guarantee UAVs' safety at intervals between discrete trajectory points. Based on convex optimization formulation, the detailed procedure of using sequential convex programming based on penalty function to solve multi-UAV trajectory planning is provided. Numerical simulations are conducted to show the effectiveness of the proposed method. The results show that the method can acquire the solution with better optimality and efficiency than the pseudospectral method, especially for larger scale UAV formation.

Key words: unmanned aerial vehicle, trajectory planning, collision avoidance, optimal control, convex programming

中图分类号: