分布式传感器网络自适应一致性融合估计算法

Adaptive Consensus Fusion Estimation Algorithm for Distributed Sensor Networks

针对分布式传感器网络的目标一致性状态估计问题,提出自适应一致性融合估计算法。考虑到网络中节点为测距和测方位的传感器,基于观测噪声与目标状态相关的假设,构建量测模型;引用无迹卡尔曼滤波与CI算法得到各节点的局部估计,通过误差矩阵加权更新节点状态以改进一致性算法,实现各节点对目标状态的一致性估计。仿真实验结果表明,该算法能够在快速收敛的过程中实现无中心节点的分布式传感器网络中各节点对目标位置的精确估计,同时又保证各节点之间的一致性。

To get the consensus state estimation about the target based on distributed sensor networks,the adaptive consensus fusion estimation algorithm is proposed. In this paper,nodes in distributed sensor networks are considered as ranging and bearing sensors. Based on the assumption about the measured noise and the target state,the measurement model is established. Using untracebale Kalman filter and CI algorithm,the local estimation of each node is derived. Last,the improved consensus technique is employed to derive an implementation of consensus state estimation. The simulation results show that the proposed method can significantly improve the accuracy of target state estimation of each sensor node in a distributed sensor network,while the distributed consensus between each node is assured.