基于EMD-DESN的无人机集群航迹目的地预测

Destination prediction of UAV cluster trajectory based on EMD-DESN

无人机(unmanned aerial vehicle,UAV)集群作战样式多样、运动模式复杂,导致集群航迹目的地难以预测。为解决上述问题,本文提出了一种基于经验模态分解(empirical mode decomposition,EMD)和深度回声状态网络(deep echo state network,DESN)的UAV集群航迹目的地预测算法。为使集群运动模型更真实地模拟UAV集群作战过程,本文引入航向误差时变方差,改进了Olfati-Saber集群运动模型的虚拟领导项。为处理因群内的协同作用和集群航向误差导致的运动非平稳性,引入了EMD,对UAV航迹序列进行重构。考虑到获知航迹的时序性,设计了滑窗结构,采用DESN对重构航迹的不同时段进行目的地预测。仿真实验结果表明,本文提出的EMD-DESN算法较基本DESN算法能以更高的准确度预测UAV集群航迹目的地,并能更早地实现稳定的正确预测。

The unmanned aerial vehicle(UAV)cluster has various combat styles and complex movement modes,which makes it difficult to predict the cluster trajectory destination.To solve the above problems,we propose a UAV cluster trajectory destination prediction algorithm based on empirical mode decomposition(EMD)and deep echo state network(DESN)in this paper.In order to make the cluster movement model more realistic to simulate the UAV cluster combat process,we introduce the time-varying variance of heading error,and improves the virtual leadership term of Olfati-Saber cluster movement model.In order to deal with the non-stationary motion caused by the coordination within the cluster and the cluster heading error,EMD is introduced to reconstruct the UAV trajectory sequence.Considering the timing of the acquired trajectory,the sliding window structure is designed,and the DESN is used to predict the destination in different periods of the reconstructed trajectory.The simulation results show that the EMD-DESN algorithm proposed in this paper can predict the UAV cluster trajectory destination with higher accuracy than the basic DESN algorithm,and can achieve stable and correct prediction earlier.