项目名称: 基于一致性的分布式滤波及其在无线网时钟同步中的应用
项目编号: No.61273110
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 田玉平
作者单位: 东南大学
项目金额: 80万元
中文摘要: 分布式滤波是传感器网络信息融合和时钟同步的重要基础。本项目重点研究在有通信延时和数据丢包情况下基于一致性的线性分布式滤波算法的设计、收敛性分析以及在时钟同步中的应用。研究计划主要包括:(1) 采用一致性协议和线性观测器相结合的方法,设计分布式滤波的基础算法;(2) 用频域凸分析和图论为工具分析基础滤波系统在有非均匀通信延时情况下的稳定性,并通过引入自时延对滤波算法进行修正;(3)用时频转换、马尔科夫跳变系统稳定性理论和线性矩阵不等式方法设计分布式递推滤波算法的增益矩阵,并给出算法在测量噪声环境下和丢包网络中的均方收敛的条件;(4)以传感器网时钟同步问题为应用对象在OPNET仿真平台上对算法进行仿真实验研究。
中文关键词: 分布式滤波;一致性;时钟同步;多智能体系统;传感器网络
英文摘要: Distributed filtering serves as an important basis of information fusion and clock synchronization in sensor networks. This project focuses on the algorithm design, convergence analysis of consensus-based linear distributed filtering under communication delays and packet losses, and its application to the problem of clock synchronization. The research plan mainly includes the following: (1) design of basic algorithms of distributed filtering by combining the consensus protocol and the linear observer; (2) stability analysis of the basic filtering system with diverse communication delays by using the frequency-domain method and graph theory, and modification of the basic algorithm via introducing self-delays; (3) design of gain matrices of distributed recursive filtering algorithms by using the techniques of transformation from the frequency-domain to the time-domain, stability theory for Markov jumping systems and linear matrix inequalities, and mean-square convergence analysis of the recursive algorithms under measurement noises and packet losses; (4) simulation experiment research on the problem of clock synchronization by implementing the designed algorithms using the software OPNET.
英文关键词: distributive filtering;consensus;clock synchronization;multi-agent system;sensor network