基于自适应加权融合的分布式滤波算法

针对存在丢包的传感器网络中每个传感器节点对目标估计确信度不同的问题,提出一种基于自适应加权融合的分布式滤波算法.考虑节点在网络中的影响力及其节点属性,将节点重要度与传感器网络节点观测数据间的支持度线性加权,获得每个传感器节点对目标的估计确信度,并将该确信度构成的融合权值引入节点状态估计值的一致性协议中,更新传感器节点对目标的状态估计值,提高分布式滤波算法的估计精度和传感器节点估计值的一致性.仿真结果验证了所提出方法的有效性.

Adaptive weighted fusion based linear distributed filtering algorithm

Aiming at the problem that each node has different certainty degree to the target in sensor network with packetdropping, an adaptive weighted fusion based linear distributed filtering algorithm is proposed. Considering the node property and the node influence over the network, node importance and the mutual support degree from observation data are fused by linear weights to attain the certainty degree, and then the fused weights composed by the certainty degree are introduced into the consensus protocol to update the state estimate at each node for the target. The proposed algorithm can not only improve the accuracy of the distributed filtering, but also enhance the consistency of the state estimates. Simulation results demonstrate the effectiveness of the proposed algorithm.