[1]吴 莎,杨小军.基于 Cauchy-Schwarz 散度的多传感器控制[J].计算机技术与发展,2020,30(06):160-166.[doi:10. 3969 / j. issn. 1673-629X. 2020. 06. 031]
　WU Sha,YANG Xiao-jun.Multi-sensor Control Based on Cauchy-Schwarz Divergence[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2020,30(06):160-166.[doi:10. 3969 / j. issn. 1673-629X. 2020. 06. 031]
点击复制

基于 Cauchy-Schwarz 散度的多传感器控制()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
30
期数:
2020年06期
页码:
160-166
栏目:
应用开发研究
出版日期:
2020-06-10

文章信息/Info

Title:
Multi-sensor Control Based on Cauchy-Schwarz Divergence
文章编号:
1673-629X(2020)06-0160-07
作者:
吴 莎; 杨小军
长安大学 信息工程学院,陕西 西安 710064
Author(s):
WU Sha; YANG Xiao-jun
School of Information Engineering,Chang’an University,Xi’an 710064,China
关键词:
多目标跟踪; 传感器控制; LMB 滤波器; 柯西-施瓦兹准则; 鲁棒广义协方差交叉准则
Keywords:
multi-target tracking; sensor control; LMB filter; Cauchy-Schwarz criterion; R-GCI rule
分类号:
TP273
DOI:
10. 3969 / j. issn. 1673-629X. 2020. 06. 031
摘要:
对传感器网络下的多目标跟踪问题,基于 Cauchy-Schwarz 准则提出一种多传感器多目标跟踪的传感器控制方法,主要包括多目标跟踪、信息融合和传感器决策三大部分。 将信息理论的柯西-施瓦兹(CS) 散度作为多传感器控制的目标函数,利用目标函数对传感器所采取的决策的优劣进行评价。 采用带标签多伯努利滤波器(LMB) ,基于鲁棒广义协方差交叉(R-GCI) 准则对多目标概率密度进行分布式融合,获得多目标运动的航迹估计。 基于序贯蒙特卡洛(SMC) 方法得到CS 散度和 LMB 滤波器的逼近执行,对传感器决策的选择问题,由于最优算法在实际应用中存在计算难且代价高的问题,故而采用局部搜索算法得到多传感器控制问题的次优解。 仿真结果表明了基于 CS 散度的多目标跟踪多传感器控制算法的可行性和优越性。
Abstract:
For multi-target tracking in sensor networks,based on Cauchy-Schwarz criterion, we propose a multi-sensor multi-target tracking sensor control method which mainly includes multi-target tracking,information fusion and sensor decision-making. The Cauchy-Schwarz (CS) divergence of information theory is taken as multi-sensor control objective function to evaluate the decision made by the sensor. Based on the robust generali-? zed covariance intersection (R-GCI) rules to the distributed fusion of multiple objective probability density,track estimation of multi-target motion is obtained by using labeled Bernoulli filter (LMB) . Based on sequential Monte Carlo (SMC) method,the approximate execution of CS divergence? and LMB filter is obtained. As the optimal algorithm is difficult and costly in practical application,a local search algorithm is proposed to obtain the subo-ptimal solution of multi - sensor control problem. The simulation results show the feasibility and superiority of the multi-target tracking and multi-sensor control algorithm based on CS divergence.

相似文献/References:

[1]石亚玲,刘正熙,熊运余,等.基于弱特征重识别的多目标长效摘要[J].计算机技术与发展,2018,28(05):27.[doi:10.3969/j.issn.1673－629X.2018.05.007]
　SHI Ya-ling,LIU Zheng-xi,XIONG Yun-yu,et al.Long Term Abstraction of Multi－object Based on Weak Feature Re－identification[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2018,28(06):27.[doi:10.3969/j.issn.1673－629X.2018.05.007]
[2]季露,陈志,岳文静.基于模型融合和特征关联的视频目标跟踪算法[J].计算机技术与发展,2018,28(06):93.[doi:10.3969/ j. issn.1673-629X.2018.06.021]
　JI Lu,CHEN Zhi,YUE Wen-jing.Video Target Tracking Algorithm Based on Model Fusion and Feature Association[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2018,28(06):93.[doi:10.3969/ j. issn.1673-629X.2018.06.021]

更新日期/Last Update: 2020-06-10