非均匀拓扑网络中的分布式一致性状态估计算法

Distributed consensus state estimation algorithm in asymmetrical networks

分布式一致性状态估计是传感器网络中节点对目标的一种有效的估计融合方法。针对网络非均匀拓扑情况下的一致性状态估计问题,首先,研究了分布式传感器网络一致性状态估计框架,提出了四级功能模型,从信息处理、交互及融合的角度描述了一致性状态估计技术的主要流程;其次,考虑网络非均匀拓扑时一致性收敛速度较慢的情况,根据节点间通信链接的重要性设计了基于动态拓扑信息的自适应权值分配方法,在此基础上提出了基于自适应加权的卡尔曼一致性滤波(adaptive weighted Kalman consensus filter,AW-KCF)算法。仿真结果显示,AW-KCF在非均匀拓扑的稀疏网络中具有较快的一致性收敛速度。

Distributed consensus state estimation（DCSE）is an effective fusion method in sensor networks.For the DCSE problem with asymmetrical topology,a four-layer functional model which describes the main flow of consistency state estimation technology is firstly proposed from the perspective of information processing,interaction and fusion.And then,considering the network topology of asymmetrical consistency of slow convergence rate,an adaptive assignment algorithm for consensus-rate factor based on dynamic topology information is designed according to the importance of communication links between nodes,and a Kalman consensus filter named adaptive weighted Kalman consensus filter（AW-KCF）is proposed based on the adaptive consistency rate factor.Finally,the simulation results indicate that AW-KCF outperforms KCF with faster convergence rate in sparse networks with asymmetrical topology.