基于Fisher信息距离的传感器管理方法

在进行面向跟踪的传感器管理方法研究时，传统的思路主要以跟踪精度为导向，但是在数据关联环节，改善或维持目标之间的区分度才是传感器资源分配的主要目的。需求的偏差导致不合理的传感器资源分配结果。依据信息几何理论，以流形中的Fisher信息距离来定义目标之间的区分度，然后直接依据目标之间的区分度制定传感器资源分配策略。关联仿真实验表明，该方法相较于传统的传感器管理方法，有效提高了目标之间的关联正确率。

Method for sensor management based on Fisher information distance

While facing the research of tracking-oriented sensor management,the traditional way focuses on the precision of tracking.But the essential demand of the sensor resources is to improve or maintain the resolvability among different targets at the segment of data association,and the bias of demand leads to unreasonable distribution of sensor resources.According to the theory of information geometry,the resolvability defined by Fisher information distance in manifold is proposed.Then the distribution strategy of sensor resources is established based on the resolvability among targets directly.Compared with the traditional way,the simulation results show that the proposed method effectively improves the valid association rates.