引用本文:梁宵,王宏伦,骆海涛.复杂三维多面体环境中空地协作追逃问题[J].控制理论与应用,2021,38(5):623~633.[点击复制]
LIANG Xiao,WANG Hong-lun,LUO Hai-tao.Collaborative pursuit-evasion of air-ground system in a complex 3D polyhedral map[J].Control Theory and Technology,2021,38(5):623~633.[点击复制]
复杂三维多面体环境中空地协作追逃问题
Collaborative pursuit-evasion of air-ground system in a complex 3D polyhedral map
摘要点击 1748  全文点击 590  投稿时间：2019-10-29  修订日期：2020-09-13
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2020.90896
  2021,38(5):623-633
中文关键词  追逃博弈  空地协作系统  复杂三维多面体环境  走法生成器  最坏情况
英文关键词  pursuit-evasion game  air-ground system  complex 3D polyhedral map  move generator  worst case
基金项目  国家自然科学基金项目(61973222, 61503255), 辽宁省自然科学基金项目(2019–ZD–0247), 辽宁省“兴辽英才计划”项目(XLYC1907179)资助.
作者单位邮编
梁宵* 沈阳航空航天大学 自动化学院 110136
王宏伦 北京航空航天大学 自动化科学与电气工程学院 100191
骆海涛 中国科学院 沈阳自动化研究所 110016
中文摘要
      结合无人机(UAV)的空中移动和无人车(UGV)的地面移动特点, 本文提出了一种UAV/UGV空地协作系统, 并且针对其在复杂地形中的追逃问题, 提出了一种复杂三维多面体环境中UAV/UGV空地协作追逃策略. 首先介绍 了UAV/UGV空地协作系统的结构与协作追逃问题描述. 接着将边界值问题(BVP)改进并离散化作为博弈走法生成 器. 然后, 针对逃方已知追方位置, 而追方只具备直线视野(LOS)的不利条件, 分析了最坏情况. 逃方策略在保证最 大生存条件下尽可能获得博弈胜利. 追方策略根据逃方状态分成3种情况进行讨论: 逃方处于追方的视野范围内、 逃方刚刚消失于追方视野以及追方完全丢失逃方的情况. 最后, 对比仿真结果说明了本文算法的有效性, 并分析了 追逃结果的影响因素. 由于地形是非凸的并且充满障碍, 因此该策略虽不能保证追方一定能够胜利, 但在最坏情况 下是最优的.
英文摘要
      Combining the characteristic of unmanned aerial vehicle (UAV) and unmanned ground vehicle (UGV), a UAV/UGV air-ground system and its collaborative pursuit-evasion strategy in a complex 3D polyhedral map is proposed. The structure of the UAV/UGV air-ground system and the pursuit-evasion game is introduced first. Then, the discrete boundary value problem (BVP) is improved and used as move generator. According to the case of that evader knows the position of pursuers at any time but pursuers just have a line-of-sight (LOS) view, the worst case is analyzed for both of players. In the game, evader will try to win the game under the premise of ensuring survival, and the strategy of pursuers is discussed in three situations: one is evader is in the sight of pursuers, one is the position of evader is known by pursuers before a while and the last one is the position of evader is completely unknown to pursuers. The contrastive simulation results show that the method is effective and optimal in the worst case and the influencing factors of the pursuit-evasion result is also analyzed.