基于分布式有限感知网络的多伯努利目标跟踪

针对感知范围受限的分布式传感网多目标跟踪问题, 在多伯努利滤波跟踪理论基础上提出分布式视场互补多伯努利关联算术平均融合跟踪方法. 首先, 通过视场互补扩大传感器感知范围, 其中, 局部公共区域只互补一次以降低计算成本. 其次, 每个传感器分别运行局部多伯努利滤波器, 并将滤波后验结果与相邻传感器进行泛洪通信使得每个传感器获取多个相邻传感器的后验信息. 随后, 通过距离划分进行多伯努利关联, 将对应于同一目标的伯努利分量关联到同一个子集中, 并对每个关联子集进行算术平均融合完成融合状态估计. 仿真实验表明, 所提方法在有限感知范围的分布式传感器网络中能有效地进行多目标跟踪.

Multi-Bernoulli Target Tracking Based on Distributed Limited Sensing Network

In order to solve the problem of multi-target tracking in distributed sensor networks with limited sensing range (LSR), a distributed arithmetic average (AA) fusion multi-Bernoulli filter is proposed based on field of view complementation and multi-Bernoulli association. First, the sensor＇s sensing range is expanded by complementing the field of view, in which the local common areas are complemented only once to reduce the calculation cost. Secondly, each sensor separately operates a local multi-Bernoulli filter and conducts flood communication between neighbor sensors over the filter posteriors, so that each sensor can obtain posterior information of multiple sensors. Then, multi-Bernoulli correlation is performed by distance division to associate Bernoulli components corresponding to the same target to the same subset, and the AA fusion is performed for each associated subset to complete fusion state estimation. Simulation results show that the proposed method can effectively track multiple targets in distributed sensor networks with LSR.