摘要
针对传统无线传感器网络(Wireless Sensor Network,WSN)对运动目标的定位和跟踪容易产生明显误差的问题,提出利用改进FOA-GRNN和迭代Cubature卡尔曼滤波的实时目标跟踪方法。基于改进FOA-GRNN法,利用从锚点接收到的运动目标的模拟(RSSI)值和相应的实际目标二维位置对GRNN进行训练,从而获得单个目标在二维运动时的准确初始位置;利用迭代Cubature卡尔曼滤波法对实时目标进行精准定位和测距,获得实时目标的准确定位和跟踪信息;将改进的FOA-GRNN法和迭代Cubature卡尔曼滤波法相结合用于WSN中实时目标跟踪和定位,在提高初始位置精度的同时,还提高了实时目标定位和跟踪信息的准确度。实验结果表明,相比其他几种较新的方法,该方法改善了WSN中实时目标的跟踪性能,降低了误差,提高了跟踪精度。
Aiming at the problem that the traditional wireless sensor network(WSN)produces obvious errors in the positioning and tracking of moving targets,this paper proposes a real-time target tracking method by using improved fly optimization algorithm general regression neural network(FOA-GRNN)and iterative Cubature Kalman filtering.Based on the improved FOA-GRNN method,GRNN was trained by using the simulated value(RSSI)of the moving target received from the anchor and using the corresponding two-dimensional position of the actual target,so as to obtain the accurate initial position of a single target in two-dimensional motion.We adopted iterative Cubature Kalman filtering method for accurate positioning and ranging of real-time targets,so as to get accurate positioning and tracking information of real-time targets.The improved FOA-GRNN method and iterative Cubature Kalman filtering method were combined for real-time target tracking and positioning in WSN,which ensured the initial position accuracy,and improved the accuracy of real-time target positioning and tracking information as well.The experiments show that compared with existing methods,the proposed method enhances the tracking performance of real-time targets in WSN,reduces errors and improves tracking accuracy.
作者
罗宏等
蓝耿
聂良刚
粟光旺
伍一坤
Luo Hongdeng;Lan Geng;Nie Lianggang;Su Guangwang;Wu Yikun(Guangxi University of Finance and Economics,Nanning 530003,Guangxi,China;Guangxi University,Nanning 530001,Guangxi,China)
出处
《计算机应用与软件》
北大核心
2021年第12期135-141,219,共8页
Computer Applications and Software
基金
广西科技厅重点研发计划项目(2017AB18048)
广西自然科学基金联合资助培育项目(2018GXNSFAA294010)。
关键词
卡尔曼滤波
无线传感器网络
改进的FOA-GRNN
迭代Cubature
实时目标跟踪
Kalman filter
Wireless sensor network(WSN)
Improved fly optimization algorithm general regression neural network(FOA-GRNN)
Iterative Cubature
Real-time target tracking
