无线传感器网络多源数据融合技术研究
发布时间：2018-11-24 17:12
【摘要】：摘要：随着无线传感器网络技术的飞速发展和广泛应用,传感器网络内的感知数据大量增加,如何实现传感网中的数据融合,是目前无线传感器网络领域内的热点研究课题之一。在复杂的无线传感器网络中,由于传感器节点能量受限,同时传感网中通信数据量大、数据种类繁多,因此需要对多源的无线传感器网络数据进行融合处理,从而减少数据通信量,提高数据准确度,延长无线传感器网络的生命周期。 本文在分析多源数据融合技术的相关理论基础上,重点研究了Dempster-Shafer证据理论(简称DS证据理论)在无线传感器网络多源数据融合过程中的若干关键问题,主要包括以下几个方面： (1)基于DS证据理论,提出了一种可应用于无线传感器网络实际应用的信度分配策略。考虑到已知样本数据的分布特点对于信度分配结果的合理影响,本文提出的证据理论信度分配策略采用马氏距离作为被测目标与样本数据间的距离计算方法,并在考虑到闭合世界假设和开放世界假设的情况下,构建基于指数函数形式的基本信度分配函数,同时设计了相应的证据理论信度分配算法。 (2)引入向量空间的概念,提出了基于余弦定理的证据冲突表示方法。本文在分析了现有的几类典型数据冲突程度表示方法后,指出了各自的特点和存在的不足。考虑到计算复杂度的影响,本文应用Pignistic概率函数进行证据信度分配向量的降维,并以向量空间中的角度概念衡量证据间的冲突程度。 (3)为了解决冲突数据的融合问题,本文借鉴证据加权平均思想,改进了一种基于证据间支持程度的数据融合算法。文中应用余弦定理得到证据间的相互支持程度,构建证据相关矩阵,得到各个证据的可信度权值,再应用Murphy方法完成数据融合。 论文最后根据轨道交通安全运营需求,提出并设计实现了基于证据理论多源数据融合策略的轨道交通变压器故障诊断系统,并对涉及到的功能架构设计、数据库设计、各核心功能模块设计进行了介绍,最后选取实例对系统进行了功能验证,取得了良好了效果。
[Abstract]:Absrtact: with the rapid development and wide application of wireless sensor network (WSN) technology, the perceptual data in wireless sensor network (WSN) is increasing greatly, so how to realize the data fusion in WSN. It is one of the hot research topics in the field of wireless sensor networks. In the complex wireless sensor networks, because of the limited energy of sensor nodes, the large amount of communication data and the variety of data in the sensor network, it is necessary to fuse the multi-source wireless sensor network data. In order to reduce the amount of data communication, improve the accuracy of data, extend the life cycle of wireless sensor networks. Based on the analysis of the theory of multi-source data fusion, this paper focuses on the key issues of Dempster-Shafer evidence theory (DS evidence theory) in the process of multi-source data fusion in wireless sensor networks. The main contents are as follows: (1) based on DS evidence theory, a reliability allocation strategy is proposed for practical application in wireless sensor networks (WSN). Considering the reasonable influence of the distribution characteristics of the known sample data on the reliability distribution results, the evidence theory reliability allocation strategy in this paper uses Markov distance as the distance calculation method between the measured target and the sample data. Considering the closed world hypothesis and the open world hypothesis, the basic reliability assignment function based on exponential function is constructed, and the corresponding evidence theory reliability assignment algorithm is designed. (2) introducing the concept of vector space, a method of evidence conflict representation based on cosine theorem is proposed. After analyzing several kinds of typical data conflict degree representation methods, this paper points out their characteristics and shortcomings. Considering the influence of computational complexity, this paper applies the Pignistic probability function to reduce the dimension of the evidence reliability assignment vector, and measures the degree of conflict between the evidence by the angle concept in the vector space. (3) in order to solve the problem of conflict data fusion, this paper improves a data fusion algorithm based on the support degree of evidence by using the idea of weighted average of evidence. In this paper, the degree of mutual support of evidence is obtained by using cosine theorem, and the correlation matrix of evidence is constructed, and the credibility weight of each evidence is obtained, and then the data fusion is completed by using Murphy method. Finally, according to the requirements of rail transit safety operation, a fault diagnosis system of transformer based on multi-source data fusion strategy based on evidence theory is proposed and implemented. The design of each core function module is introduced. Finally, the system is verified by an example, and a good result is obtained.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP212.9;TN929.5
本文编号：2354438
[Abstract]:Absrtact: with the rapid development and wide application of wireless sensor network (WSN) technology, the perceptual data in wireless sensor network (WSN) is increasing greatly, so how to realize the data fusion in WSN. It is one of the hot research topics in the field of wireless sensor networks. In the complex wireless sensor networks, because of the limited energy of sensor nodes, the large amount of communication data and the variety of data in the sensor network, it is necessary to fuse the multi-source wireless sensor network data. In order to reduce the amount of data communication, improve the accuracy of data, extend the life cycle of wireless sensor networks. Based on the analysis of the theory of multi-source data fusion, this paper focuses on the key issues of Dempster-Shafer evidence theory (DS evidence theory) in the process of multi-source data fusion in wireless sensor networks. The main contents are as follows: (1) based on DS evidence theory, a reliability allocation strategy is proposed for practical application in wireless sensor networks (WSN). Considering the reasonable influence of the distribution characteristics of the known sample data on the reliability distribution results, the evidence theory reliability allocation strategy in this paper uses Markov distance as the distance calculation method between the measured target and the sample data. Considering the closed world hypothesis and the open world hypothesis, the basic reliability assignment function based on exponential function is constructed, and the corresponding evidence theory reliability assignment algorithm is designed. (2) introducing the concept of vector space, a method of evidence conflict representation based on cosine theorem is proposed. After analyzing several kinds of typical data conflict degree representation methods, this paper points out their characteristics and shortcomings. Considering the influence of computational complexity, this paper applies the Pignistic probability function to reduce the dimension of the evidence reliability assignment vector, and measures the degree of conflict between the evidence by the angle concept in the vector space. (3) in order to solve the problem of conflict data fusion, this paper improves a data fusion algorithm based on the support degree of evidence by using the idea of weighted average of evidence. In this paper, the degree of mutual support of evidence is obtained by using cosine theorem, and the correlation matrix of evidence is constructed, and the credibility weight of each evidence is obtained, and then the data fusion is completed by using Murphy method. Finally, according to the requirements of rail transit safety operation, a fault diagnosis system of transformer based on multi-source data fusion strategy based on evidence theory is proposed and implemented. The design of each core function module is introduced. Finally, the system is verified by an example, and a good result is obtained.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP212.9;TN929.5
【参考文献】
相关期刊论文 前7条
1 刘宗香;谢维信;黄敬雄;;一种新的基于概率理论的概率数据互联滤波器[J];电子与信息学报;2009年07期
2 熊伟;张晶炜;何友;;基于多维分配和灰色理论的航迹关联算法[J];电子与信息学报;2010年04期
3 陈羽中;陈亦萍;;无线传感器网络数据融合技术的研究进展[J];福建电脑;2009年11期
4 张轮;瞿军;陆琰;陈晨;;城市轨道交通的无线传感器网络模型[J];城市轨道交通研究;2007年12期
5 何友;薛培信;王国宏;;一种新的信息融合功能模型[J];海军航空工程学院学报;2008年03期
6 赵宗贵;;信息融合技术现状、概念与结构模型[J];中国电子科学研究院学报;2006年04期
7 韩峰;朱镭;智小军;;基于模糊理论的多传感器数据融合测量[J];应用光学;2009年06期
，本文编号：2354438
本文链接：https://www.wllwen.com/kejilunwen/wltx/2354438.html
