时变定位误差的分布式雷达组网航迹融合算法

Track fusion algorithm for distributed radar network based time-varying location error

为解决相关分布式卡尔曼滤波（DDKF）航迹融合算法在应用于时变定位误差的分布式雷达组网中,各接收站点之间需要建立通信拓扑架构传输时变定位误差的局限性,提出了改进的相关分布式卡尔曼滤波算法（IDDKF）。通过各接收站点根据自身定位误差构造解相关的局部航迹状态信息,融合中心在融合时刻重构航迹状态的全局后验概率,由此得到全局最优的航迹状态估计量。基于状态向量协方差矩阵的分析表明,IDDKF算法的融合性能明显优于传统的最大似然算法,在避免建立通信拓扑架构的前提下,达到了DDKF算法的融合性能。

When the decorrelated distributed Kalman filter algorithm is applied to distributed radar network with timevarying location error,it is necessary to build communication infrastructure to transfer time-varying location error among receiving stations.To solve this limitation,this paper presents the improved decorrelated distributed Kalman filter algorithm（IDDKF）.After each receiving station constructs decorrelated local track information based on its location error,the fusion center reconstructs global posterior possibility at fusion instant.Hence,the IDDKF algorithm can provide the global optimal estimator.Simulation analysis based on the covariance matrix of the state vector verifies that the IDDKF provides more precise fusion track than the traditional maximum likelihood algorithm,and the performance is comparable with that of the DDKF while the communication infrastructure is avoided being built.