基于运动目标预测的多无人机分布式协同搜索

摘要

Cooperative search of multi-unmanned aerial vehicle (multi-UAV)is an important research field of multi-UAV cooperative control.With the complication of battlefield environment,it is especially important for UAVs to search for dynamic time-critical targets.For motion characteristics of dynamic time-critical targets, contributing the motion prediction model of dynamic time-critical targets first of all,in order to decrease the un-certainty of targets motion and optimize search performance of UAVs,putting forward a transition probability density of targets based on gauss distributing on the basis of targets existence probability which is updated gby Bayesian theory,and computing existence probability of dynamic time-critical targets,then,contributing char-acteristics index function of UAVs cooperative search with reason,in the frame of distributed model prediction control (DMPC),transforming the centralized on-line optimization problem of multi-UAV to a distributed on-line optimization problem of each UAV,finally,solving the characteristics index function optimally.Simulation results demonstrate the validity of the proposed method.%多无人机（unmanned aerial vehicle，UAV）协同搜索是多UAV协同一个重要研究方向。随着战场环境的复杂化，UAV对动态时敏目标的搜索显得尤为重要。针对动态时敏目标的运动特性，首先建立动态时敏目标的运动预测模型，为了降低由于目标运动造成的不确定性以及优化UAV的搜索性能，在采用贝叶斯理论对目标存在概率进行更新的基础上，研究了一种基于高斯分布的目标转移概率密度，通过计算得到动态时敏目标的存在预测概率，然后合理建立UAV协同搜索的性能指标函数，在分布式模型预测控制框架下，将多UAV集中式在线优化问题转化为各架UAV的分布式在线优化问题，最后对性能指标函数进行优化求解。通过仿真验证了所提方法的有效性。