随机网格回归Monte Carlo双UAVs最优目标协调跟踪

Stochastic Mesh Regression Monte Carlo Based UAVs Optimal Target Tracking

针对多个无人机(unmanned aerial vehicle,UAV)执行基于视觉的目标跟踪的最佳协调问题,提高不可预知的地面目标的最佳结合点的视觉测量效果,提出了一种基于随机网格回归Monte Carlo的UAV最优目标跟踪策略。首先,通过无人机动力学和目标动力学分析,获得双UAV情况下的随机最优协调控制目标;其次,针对提出的控制目标,引入Monte Carlo求解方案,同时为解决标准Monte Carlo方案中存在的状态空间维度较高,计算复杂且精度不高的问题,利用随机网格方式构建回归Monte Carlo方案,实现UAV的最优协调控制;最后,通过仿真实验验证了所提方法的有效性。

In order to solve the optimal coordination problem of visual target tracking for unmanned aerial vehicles (UAVs), and improve the visual measurement result in the best combination of unpredictable target on the ground, this paper presents the stochastic mesh regression Monte Carlo based UAVs optimal target tracking algorithm. Firstly, the stochastic optimal coordinated control target is obtained by the analysis of UAV dynamics and target dynamics. Secondly,for the control objectives, the Monte Carlo method is introduced, and the random grid based regression Monte Carlo scheme is also designed to solve the high state space dimension of the standard Monte Carlo scheme, so as to solve the complex calculation and low precision, which can realize UAVs optimal coordinated control. Finally, simulation experiments are carried out to verify the effectiveness of the proposed method.