基于模糊神经网络的分布式多传感多目标信息融合

针对分布式融合系统存在计算组合爆炸、难以推广到多站和多层感知器模型学习速度慢且容易收敛到局部极小等缺陷，提出了一种基于模糊神经网络的分布式多传感器多目标状态信息融合方法。仿真结果证明了该方法的快速性、准确性和有效性。

Fuzzy Neural-Network-Based State Information Fusion

In recent years, multisensor information fusion has received significant attention.In this paper, a new distributed multisensor multitarget state information fusion, including algorithm and structure, is studied.This new fusion tries to overcome two existing shortcomings: (1) combinatorial explosion in computation; (2) low convergence speed and failure to give globally minimum error.Fuzzy neural network is introduced; a new distributed construction of multisensor multitarget state information fusion is given schematically in Fig. 2. In Fig. 2, "A" and "B " represent estimates of sensor 1 and 2 respectively, "C" represents fuzzy neural network. Eqs. (9)through (17) give the fusion algorithm. The error norms as given in eqs. (2), (3) and (4)are particularly important for reducing noise effect.Table 1 gives the analysis of properties and Figs. (3), (4) and (5) give simulation results. They prove that the new algorithm is effective and reliable.