快速协方差交叉融合算法及应用  

A Fast Covariance Intersection Fusion Algorithm and Its Application

作　　者：从金亮 李银伢[2] 戚国庆[2] 盛安冬[2] CONG Jin-Liang;LI Yin-Ya;QI Guo-Qing;SHENG An-Dong(School of Electric and Automatic Engineering,Changshu Institute of Technology,Suzhou 215500;School of Automation,Nanjing University of Science and Technology,Nanjing 210094)

机构地区：[1]常熟理工学院电气与自动化工程学院,苏州215500 [2]南京理工大学自动化学院,南京210094

出　　处：《自动化学报》2020年第7期1433-1444,共12页Acta Automatica Sinica

基　　金：国家自然科学基金(61871221,61876024);国防基础研究项目(JCKY2018209B010);江苏省高等学校自然科学研究项目(19KJB510015)资助。

摘　　要：针对分布式传感网络系统中存在互协方差未知的情形,融合系数的科学设计对于融合性能至关重要.本文以各节点估计方差矩阵逆的迹的倒数作为计算融合系数的中间变量,设计了一种序贯快速协方差交叉融合算法,可以显著减少各个融合节点的计算量,能够保证各融合节点融合结果相同.在给定系统的误差方差上界约束与优化指标前提下,该融合算法结合粒子群优化算法,能够给出对分布式系统中各个节点的传感器精度要求.工程实践中,可为传感器的选型提供理论依据.最后,给出了一个分布式网络传感器精度选型的算例及快速协方差交叉融合算法在雷达网中的应用实例.In the distributed sensor net-system with unknown cross-covariance,the design of fusion coefficients is crucial for the fusion performance.In this paper,a fast covariance intersection algorithm is presented,which can significantly reduce the computational complexity of each fusion node and ensure that the fusion result of each node is the same.The new fusion coefficients can be calculated straightforward by taking the reciprocal of the trace of the inverse variances as local fusion coefficients and using an iterative process for fusion step to revise the co-efficient weight.For given upper bound of the fusion error variance,the proposed fusion algorithm which combines with the idea of particle swarm optimization can give the sensor precision requirements of each node for the distributed system.In the engineering practice,it can provide a theoretical basis for the selection of sensors.Finally,a real radar system example is provided to verify the effectiveness of the proposed algorithm and an application example is given for the sensor selection of distributed network.

关 键 词：分布式传感网 估计融合 协方差交叉 传感器选型 方差约束 

分 类 号：TP212.9[自动化与计算机技术—检测技术与自动化装置] TP301.6[自动化与计算机技术—控制科学与工程]