航空学报 > 2011, Vol. 32 Issue (11): 2073-2082   doi: CNKI:11-1929/V.20110707.1107.005

基于协同进化粒子群和Pareto最优解的卫星编队队形重构方法

黄海滨1,2, 马广富1, 庄宇飞1, 吕跃勇1   

  1. 1. 哈尔滨工业大学 航天学院, 黑龙江 哈尔滨 150001;
    2. 哈尔滨工业大学(威海) 信息与电气工程学院, 山东 威海 264200
  • 收稿日期:2011-01-26 修回日期:2011-05-04 出版日期:2011-11-25 发布日期:2011-11-24
  • 通讯作者: 马广富 E-mail:magf@hit.edu.cn
  • 作者简介:黄海滨(1983- ) 男,博士研究生。主要研究方向:卫星编队飞行、轨迹规划。 Tel: 0451-86413411-8606 E-mail: hhb833@gmail.com马广富(1963- ) 男,博士,教授,博士生导师。主要研究方向:航天器姿态控制、最优控制。 Tel: 0451-86413411-8606 E-mail: magf@hit.edu.cn
  • 基金资助:

    国家自然科学基金 (61004072);高等学校博士学科点专项科研基金(20102302110031);中央高校基本科研业务费专项资金(HIT.KLOF.2010016)

Satellite Formation Reconfiguration Using Co-evolutionary Particle Swarm Optimization and Pareto Optimal Solution

HUANG Haibin1,2, MA Guangfu1, ZHUANG Yufei1, LU Yueyong1   

  1. 1. School of Astronautics, Harbin Institute of Technology, Harbin 150001, China;
    2. School of Information and Electrical Engineering, Harbin Institute of Technology at Weihai, Weihai 264200, China
  • Received:2011-01-26 Revised:2011-05-04 Online:2011-11-25 Published:2011-11-24

摘要: 针对卫星编队自主队形重构问题,提出了基于协同进化粒子群优化(CPSO)和Pareto最优解的求解方法。首先,使用Legendre伪谱法(LPM)将队形重构问题离散化为非线性规划(NLP)问题;其次,根据卫星编队的特点及碰撞规避的需要,使用CPSO算法对重构问题采用既独立又集中的求解方式,避免了传统优化方法对梯度的求解;然后,使用一种深度-广度优先搜索(D-BFS)算法,能够高效地找到CPSO进化中所有Pareto最优解,提升了算法的效率。仿真结果表明,该方法快速有效,能够满足实时性的要求,使得卫星编队的自主运行成为可能。

关键词: 队形重构, 碰撞规避, 伪谱法, 协同进化粒子群优化, Pareto最优解

Abstract: This paper proposes an optimal trajectory planning method for satellite formation reconfiguration using co-evolutionary particle swarm optimization (CPSO) and Pareto optimal solution. First, the Legendre pseudospectral method (LPM) is employed to transform the reconfiguration problem into a parameter optimization nonlinear programming (NLP) problem. Next, according to the features of satellite formation and the constraints of collision avoidance, a CPSO algorithm is used to solve the reconfiguration problem separately in a centralized way to avoid the computational complexity of calculating the gradient information with traditional optimization methods. Then, a depth-breadth first search (D-BFS) algorithm is used to search all the Pareto optimal solutions needed by the CPSO, with which the entire redundant search could be avoided. Simulations show that the method could solve the reconfiguration problem in real time, and guarantee collision avoidance during the entire reconfiguration process even when the number of collocation points or number of satellites increases.

Key words: formation reconfiguration, collision avoidance, pseudospectral method, co-evolutionary particle swarm optimization, Pareto optimal solution

中图分类号: