传感器网络中的分布式融合状态估计算法研究

项目名称： 传感器网络中的分布式融合状态估计算法研究

项目编号： No.60874062

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 石油、天然气工业

项目作者： 孙书利

作者单位： 黑龙江大学

项目金额： 30万元

中文摘要： 本项目将研究适用于传感器网络的分布式状态估计算法。首先，针对传感器网络数据传输过程中存在的随机滞后和丢包现象，将建立新的数学模型描述数据传输的多随机滞后和多丢包问题。对带随机滞后或/和丢包的线性离散随机系统，应用射影理论提出最优和稳态状态估计新算法。进而，应用线性最小方差分布式加权融合估计算法，给出最优和稳态分布式信息融合状态估计新算法。同时，针对传感器网络中传感器能量和通信带宽的有限性等资源受限问题，对已有融合算法进行改进，并综合考虑数据传输的随机滞后、丢包和资源受限问题，将提出适用于传感器网络的资源受约束的分布式信息融合状态估计新算法，并进行性能分析。最后，给出新算法在传感器网络目标跟踪中的应用研究。该项研究具有重要的理论意义和实际应用价值。所提出的算法可广泛应用于目标跟踪、通信和信号处理等领域。

中文关键词： 随机滞后；丢包；分布式估计；信息融合；传感器网络

英文摘要： This project will study distributed state estimation algorithms fitted to sensor networks.At first,new mathematical models to describe multiple random delays and packet dropouts in data transmission will be presented according to the random delays and packet dropouts cases during data transmission of sensor networks.For linear discrete-time stochastic systems with random delays and/or packet dropouts,the new optimal and steady-state state estimation algorithms will be presented by applying projection theory.Then,the new optimal and steady-state distributed information fusion state estimation algorithms will be given by applying distributed weighted fusion estimation algorithms in the linear minimum variance sense.Furthermore,we will improve the existed fusion algorithms according to the limited source problems of the power of sensors and the limitation of communication bandwidth in sensor networks.And we will also present the new distributed information fusion state estimation algorithms fitted to the constrained sources of sensor networks and analyze the performances,where the cases of random delays,packet dropouts and limited sources in data transmission are considered synthetically.At last,we will apply the new algorithms to the target tracking in sensor networks.The study of this project has the important significance in theory and application value in practice.The proposed algorithms can be applied to the fields such as target tracking,communication and signal processing.

英文关键词： random delay;packet dropout;distributed estimation;information fusion;sensor networks