引用本文:王宏伦,雷玉鹏,姚鹏,刘畅.基于流体扰动计算的无人机三维无碰撞航路规划[J].控制理论与应用,2016,33(3):295~303.[点击复制]
WANG Hong-lun,LEI Yu-peng,YAO Peng,LIU Chang.Planning three-dimensional collision-avoidance path for UAV on the basis of disturbed fluid dynamical system algorithm[J].Control Theory and Technology,2016,33(3):295~303.[点击复制]
基于流体扰动计算的无人机三维无碰撞航路规划
Planning three-dimensional collision-avoidance path for UAV on the basis of disturbed fluid dynamical system algorithm
摘要点击 2905  全文点击 1911  投稿时间:2015-05-29  修订日期:2015-10-31
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2016.50465
  2016,33(3):295-303
中文关键词  无人机  流体扰动  航路规划  交互多模型算法  滚动优化
英文关键词  unmanned aerial vehicle (UAV)  disturbed fluid dynamical system  path planning  disturbance multiple model (IMM) algorithm  receding horizon optimization
基金项目  国家自然科学基金项目(61175084), 长江学者和创新团队发展计划(IRT13004)资助
作者单位E-mail
王宏伦* 北京航空航天大学 hl_wang_2002@126.com 
雷玉鹏 北京航空航天大学  
姚鹏 北京航空航天大学  
刘畅 北京航空航天大学  
中文摘要
      针对三维多架无人机(UAV)入侵的情况, 提出一种基于流体扰动计算方法与交互多模型(IMM)算法相结合 的无碰撞航路规划算法. 首先将流体扰动计算方法用于三维单机入侵情况下的无碰撞航路规划, 并简要分析该方 法的特点. 然后将改进的流体扰动计算方法运用到三维多机入侵的情况, 并采用机动目标多模型跟踪算法对机动 入侵机进行跟踪并对其未来运动态势进行预测, 再结合滚动优化思想在线调整影响系数, 在无人机性能约束要求下 对避撞航路进行实时的局部优化, 最终规划出一条三维可飞行的无碰撞航路. 仿真结果表明, 该算法能够规避多架 入侵机, 航路平滑可飞, 满足实时性要求.
英文摘要
      Considering the invasion of multiple unmanned aerial vehicles (UAVs), we develop a novel method for planning a three-dimensional collision-avoidance path on the basis of the interfered fluid dynamical system (IFDS) algorithm and the interacting multiple-model (IMM) algorithm. Firstly, we apply the IFDS algorithm to the case of single intruder, and analyze the characteristics of the algorithm. Secondly, the improved IFDS is used to obtain a three-dimensional collision-avoidance path in the case of multiple intruders, and the IMM algorithm is used to track and predict the state of multiple maneuvering intruders; and then, the receding horizon optimization method is applied to dynamically optimize the local collision -avoidance path by adjusting the affective coefficients with the performance constraint. Finally, a threedimensional collision-free and feasible path can be obtained. Simulation results indicate that the improved algorithm meets the real-time requirements and produces a smooth collision-free path in the case of multiple intruders.