论文部分内容阅读
为了提高无人机对目标的搜索捕获能力和对不确定度较高的区域的回访能力,进而提高多无人机协同搜索效率,提出了一种带信息素回访机制的多无人机协同目标搜索方法。首先,建立了包含目标存在概率地图、不确定地图和数字信息素地图在内的环境感知地图及其更新机理,使得无人机对任务区域中环境和目标信息的感知更加全面和准确,为无人机在线自主地进行搜索决策奠定基础。其次,考虑到任务子区域的可控回访需求,借鉴信息素“释放-传播-挥发”的特性,设计了基于信息素的网格回访机制,来引导无人机对目标存在可能性较大的区域或者不确定度较高的区域进行回访搜索。最后,设计了基于环境感知地图的协同搜索决策性能指标,在分布式滚动时域优化框架下,建立了多无人机协同搜索决策方法。使用蒙特卡罗方法验证了无人机数量、传感器探测靶面半径、传感器性能对搜索效率的影响。对比仿真表明,信息素回访机制能够保证较强的遍历能力和回访能力,使得无人机能够尽早搜索到更多的目标,尽快地降低整个搜索区域的不确定度。
In order to improve the ability of UAV to search and capture targets and the ability of returning to areas with high uncertainty, and to improve the efficiency of multi-UAV collaborative search, a multi-UAV collaborative target with pheromone return mechanism Search method. Firstly, an environment-aware map and its updating mechanism including target probability map, indeterminate map and digital pheromone map are established, which makes the UAV more completely and accurately perceive the environment and target information in the mission area. Human-based online search decisions independently laid the foundation. Second, taking into account the controllable return needs of the sub-region of the mission, a pheromone-based grid-return mechanism is designed to draw on the characteristics of the pheromone "release-propagation-volatility Large area or area with high uncertainty for return visit search. Finally, the performance index of collaborative search based on environment-aware map is designed. Under the framework of distributed rolling time domain optimization, a multi-UAV collaborative search decision-making method is established. The Monte Carlo method was used to verify the effect of the number of UAVs, the radius of the target surface and the performance of the sensor on the search efficiency. Comparative simulation shows that the pheromone mechanism can guarantee strong traversal ability and return capability, which makes UAV search more targets as soon as possible and reduce the uncertainty of the entire search area as soon as possible.