基于无线传感器网络的目标一致性跟踪算法研究

发布时间：2018-01-26 23:20

本文关键词： 无线传感器网络目标跟踪分布式滤波一致性量化一致性权值优化对数量化概率量化量化卡尔曼滤波　出处：《浙江大学》2014年硕士论文　论文类型：学位论文

【摘要】：基于无线传感器网络的目标跟踪是传感器网络应用研究的一个重要前沿,在军事防御、医疗健康、生态环境监测、生物习性研究、灾难预警与救援等领域均有重要应用。无线传感器网络目标跟踪的关键技术主要包括传感网络技术、目标识别与检测技术以及信号处理等技术。本文主要研究了用于无线传感器网络分布式目标跟踪的一致性算法,包括卡尔曼一致性滤波算法、信息一致性滤波算法和量化通信环境下的目标一致性跟踪算法。本文对算法的原理和实验结果作了详细描述。论文的主要工作有： 1、概述了无线传感器网络分布式目标跟踪算法,介绍了无线传感器网络的目标一致性跟踪模型、目标运动模型以及无线传感器网络的拓扑建模,通过将卡尔曼滤波与一致性策略相结合,讨论了卡尔曼一致性滤波算法,实现了目标的分布式跟踪。对卡尔曼一致性滤波目标跟踪算法进行了仿真实验加以验证。 2、针对卡尔曼一致性滤波跟踪算法精度不足,提出采用信息一致性滤波算法用于无线传感器网络的分布式目标跟踪,并从理论和实验两方面将其和卡尔曼一致性滤波算法进行了对比。进一步地,本文基于信息一致性滤波算法,提出了权值优化的信息一致性滤波算法,并将其与卡尔曼一致性滤波、信息一致性滤波等算法一起作了实验分析。 3、针对无线传感器网络频带资源有限的情况,从理论和实验详细论证了传感器网络为无向网络和有向网络时的量化一致性模型,讨论了包括对数量化一致性算法、概率量化一致性算法和均匀量化一致性算法,并基于概率量化一致性算法提出了量化卡尔曼一致性滤波算法,实现了量化通信环境下的目标分布式跟踪。同时对量化卡尔曼一致性滤波算法进行了仿真分析。
[Abstract]:Target tracking based on wireless sensor networks (WSN) is one of the most important research fields in WSN applications, such as military defense, medical health, ecological environment monitoring and biological habits. Disaster warning and rescue and other fields have important applications. The key technologies of target tracking in wireless sensor networks mainly include sensor network technology. Target recognition and detection technology and signal processing technology. This paper mainly studies the consistency algorithm for distributed target tracking in wireless sensor networks, including Kalman coherence filtering algorithm. In this paper, the principle and experimental results of the algorithm are described in detail. The main work of the thesis is as follows: 1. The distributed target tracking algorithm of wireless sensor network is summarized. The object consistency tracking model, target motion model and topology model of wireless sensor network are introduced. By combining the Kalman filter with the consistency strategy, the Kalman consistency filtering algorithm is discussed. The distributed target tracking is realized, and the Kalman consistent filtering algorithm is verified by simulation. 2. Aiming at the lack of accuracy of Kalman consistent filter tracking algorithm, an information consistency filtering algorithm is proposed for distributed target tracking in wireless sensor networks. Furthermore, based on the information consistency filtering algorithm, this paper proposes a weighted optimization information consistency filtering algorithm. Experimental analysis is made with Kalman consistent filtering, information consistency filtering and so on. 3. In view of the limited frequency band resources of wireless sensor networks, the quantization consistency model of sensor networks for undirected networks and directed networks is demonstrated in detail from theory and experiment. The quantization consistency algorithm, probabilistic quantization consistency algorithm and uniform quantization consistency algorithm are discussed. Based on probabilistic quantization consistency algorithm, a quantization Kalman consistency filtering algorithm is proposed. The distributed target tracking in the environment of quantized communication is realized, and the quantization Kalman coherence filtering algorithm is simulated and analyzed.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP212.9;TN929.5

【共引文献】