一种面向协同感知的多传感器自适应管理方法  

An Adaptive Multi-sensor Management Method for Cooperative Perception

作　　者：陆泽健[1] 刘筱[1] 秦永刚[1] 潘越[1] 郭继光[1] 黄胜[1] LU Ze-jian LIU Xiao QIN Yong-gang PAN Yue GUO Ji-guang HUANG Sheng(China Academy of Electronic and Information Technology, Beijing 100041, China)

机构地区：[1]中国电子科学研究院,北京100041

出　　处：《中国电子科学研究院学报》2017年第4期353-358,共6页Journal of China Academy of Electronics and Information Technology

基　　金：电子信息装备体系研究国防科技重点实验室基础研究项目(DXZT-JC-ZZ-2011-015)

摘　　要：针对多特征目标综合识别中的传感器动态调度问题,本文提出一种面向协同感知的多传感器自适应管理方法。该方法在离线获取目标识别先验概率分布的基础上,首先利用多个管控周期的累积探测特征组合,基于贝叶斯推理实现对未知目标的在线识别,然后计算感知信息熵增益作为管控系统反馈输入,利用遗传优化算法实现多传感器-多目标的自适应分配。仿真结果表明,与随机分配算法相比,该方法识别效率提高了近21秒,有效提升了传感器网络对未知目标的动态感知能力。An adaptive management method is proposed in this paper to address the problem of multi-sensor dispatching in the sense of multi-feature target recognition. The method firstly computes the prior probability distribution and then utilizes Bayesian inference to identify unknown targets based on detected feature combination from multiple detection periods. Perceptual information entropy gain is then calculated as the feedback of the dispatching system while genetic optimization algorithm is utilized to adaptively dispatch multi-sensor on line. Simulation results have shown that,compared with the random method,the identification efficiency can raise up almost 21 seconds via the proposed method,which effectively improves the dynamic perception capability of unknown targets.

关 键 词：识别 信息熵 自适应 传感器管理 

分 类 号：TN391[电子电信—物理电子学] TP273[自动化与计算机技术—检测技术与自动化装置]