分布式系统中H_∞滤波的最优融合

摘要

For distributed multi-sensor system, it is critical to optimally fuse the local estimates in the central processor. Distributed Kalman filtering attracts wide attention because it has many advantages of distributed estimation and can reach the performance of central estimation. The optimal fusion formulation for H_8 filtering in distributed uncertain system is presented, and the performance of the distributed estimation fusion can attain the optimality as same as that of the central estimation is proved. The simulation example shows the effectiveness of the optimal fusion.%分布式多传感器系统的一个关键问题是如何在融合中心将各分站的局部估计进行最优融合.分布式Kalman滤波因其具有分布式估计的诸多优点,并能达到中心式估计的最优性能而受到广泛关注.针对分布式不确定动态系统的H_8滤波给出了最优融合公式,并证明其同样能够达到中心式的融合性能,通过仿真实例验证了该最优融合的有效性.