水下无线传感器网络节点定位与目标跟踪关键技术

发布时间：2018-01-20 05:19

本文关键词： 水下无线传感器网络节点定位水下目标跟踪节点拓扑节点位置不确定　出处：《浙江大学》2016年博士论文　论文类型：学位论文

【摘要】：随着陆地资源的逐渐减少,人们越来越重视对海洋资源的开发和利用。基于水下无线传感器网络(Underwater Wireless Sensor Networks,UWSNs)的水下目标跟踪作为海洋开发和利用的重要组成部分,正在成为一个重要的研究热点。由于UWSNs节点为水下目标跟踪提供量测,节点的位置信息对水下目标跟踪至关重要,本文首先研究了 UWSNs节点定位问题,并在此基础上研究了UWSNs水下目标跟踪问题。本文的研究工作和主要贡献如下:首先,针对没有额外辅助设备和节点通信能力不强的网络,设计了一种自上而下的节点定位策略。通过定义节点自信指数,引入梯度法,以及设计三维空间两跳距离估计算法,在保证适当跟踪精度的前提下,提高了定位覆盖率。其次,针对UWSNs节点时钟同步难以实现的问题,提出了一种不依赖时钟同步的节点定位策略。利用本地时钟的时间差获取节点与锚节点的距离差,并设计了一种基于距离差的最小二乘定位法,将距离差信息转换成节点的位置。此外,考虑到UWSNs节点有限的计算和存储能力,研究了基于局部信息节点选择水下目标跟踪问题。为了克服全局信息节点选择的弊端,提出了一种基于局部信息的节点选择策略。为了保证UWSNs目标跟踪实时性要求,采用了带反馈的分布式卡尔曼融合策略,并设计了一种基于局部信息节点选择的分布式跟踪算法,降低了计算复杂度。然后,本文研究了考虑节点拓扑结构的水下目标跟踪问题,给出了量化量测条件下后验克拉美罗下界与节点拓扑的关系,并设计了一种基于多传感器粒子滤波的水下目标跟踪策略。基于上述研究成果,本文进一步研究了非理想信道对水下目标跟踪性能的影响,设计了一种非理想信道条件下的水下目标跟踪策略,并利用海试数据验证了该跟踪策略的有效性。最后,考虑到UWSNs节点的移动性,研究了节点位置不确定水下目标跟踪问题。利用一阶泰勒级数展开法对节点位置不确定性进行了建模,将节点位置不确定的影响近似转换成量测噪声。给出了节点位置不确定条件下后验克拉美罗下界与节点估计位置的关系,并设计了一种考虑节点位置不确定的水下目标跟踪策略,提高了水下目标跟踪精度。
[Abstract]:With the gradual decline of land resources. People pay more and more attention to the development and utilization of marine resources. Underwater Wireless Sensor Networks is based on underwater wireless sensor network. As an important part of ocean development and utilization, underwater target tracking is becoming an important research hotspot. UWSNs node provides measurement for underwater target tracking. The location information of nodes is very important for underwater target tracking. Firstly, the problem of UWSNs node location is studied in this paper. On this basis, the underwater target tracking problem of UWSNs is studied. The research work and main contributions are as follows: first, the network without additional auxiliary equipment and node communication ability is not strong. A top-down node location strategy is designed. By defining the node confidence index, introducing the gradient method, and designing a two-hop distance estimation algorithm in three-dimensional space, the tracking accuracy is ensured. The location coverage rate is improved. Secondly, aiming at the problem that clock synchronization of UWSNs node is difficult to realize. This paper proposes a node location strategy which does not depend on clock synchronization, uses the time difference of local clock to obtain the distance difference between node and anchor node, and designs a least-square location method based on distance difference. The distance difference information is converted to the location of the node. In addition, the limited computing and storage capacity of the UWSNs node is considered. The problem of underwater target tracking based on local information node selection is studied in order to overcome the disadvantages of global information node selection. A node selection strategy based on local information is proposed. In order to meet the real-time requirements of UWSNs target tracking, a distributed Kalman fusion strategy with feedback is adopted. A distributed tracking algorithm based on local information node selection is designed to reduce the computational complexity. Secondly, the underwater target tracking problem considering node topology is studied in this paper. The relationship between the lower bound of posterior Clemero and node topology under the condition of quantization measurement is given, and an underwater target tracking strategy based on multi-sensor particle filter is designed. In this paper, the influence of non-ideal channel on underwater target tracking performance is further studied, and an underwater target tracking strategy is designed under the condition of non-ideal channel. The effectiveness of the tracking strategy is verified by sea trial data. Finally, the mobility of UWSNs nodes is considered. In this paper, the problem of underwater target tracking with uncertain nodal position is studied, and the first order Taylor series expansion method is used to model the uncertainty of node position. The influence of node location uncertainty is approximately converted into measurement noise, and the relationship between the posteriori Clamero lower bound and the node estimation position under the condition of node position uncertainty is given. An underwater target tracking strategy considering the uncertainty of node position is designed to improve the accuracy of underwater target tracking.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP212.9;TN929.3

【参考文献】