双属性概率图优化的无人机集群协同目标搜索

为了提高不确定环境下无人机（unmanned aerial vehicle，UAV）对目标捕获能力，进而提高多UAV协同搜索效率，提出了基于双属性概率图结合改进的协同进化遗传算法（improved co-evolutionary genetic algorithm，ICEGA）的多UAV协同目标搜索方法。首先，根据环境的先验信息，在原概率图基础上引入标志位，建立基于双属性矩阵的待搜索环境概率模型，提高环境和目标的信息感知准确度；其次，定义UAV的飞行规则并结合目标先验概率图信息，建立UAV运动模型及确定最大收益的目标函数；最后，建立分布式UAV之间的信息交互模型，运用ICEGA算法优化产生最优协同决策输入航向角集合，在线实时滚动优化产生最优协同路径。实验结果表明，基于双属性概率图结合ICEGA算法更能够保证最优路径的产生，使得UAV能够准确地搜索到目标；同时，对比仿真验证了ICEGA算法能够提高UAV之间的协同性，保证了路径可行性及提高了目标搜索效率。

Cooperative searching for the multi-UAVs based on dual-attribute probability model optimization

To improve the target capture capability of unmanned aerial vehicle (UAV) in uncertain environments and the efficiency of multi-UAVs collaborative searching, the proposed method named improved co-evolutionary genetic algorithm (ICEGA) based on dual attribute probability maps is proposed for multi-UAVs cooperative target capturing. Firstly, according to the prior information of the environment, the flag bit is introduced to the original probability map, and the proposed environment probability model is built. Secondly, the flight rules of UAVs and the environment probability model are used together to establish the UAV motion model and the objective function. Finally, the distributed model between the UAVs is established, and the ICEGA algorithm is used to optimize the input heading angles set with cooperative decision-making. The experimental results show that the dual-attribute probability map working with the ICEGA algorithm can guarantee the optimal searching path, so that the UAV can accurately capture the target. At the same time, the comparable experiments prove that the ICEGA algorithm can improve the coordination between UAVs, make the flying path feasible and improve the efficiency of target capturing.