基于无线传感器网络的高速移动节点定位研究

发布时间：2019-04-26 03:55

【摘要】：无线传感器网络(Wireless Sensor Network,WSN)综合了嵌入式计算技术、传感器技术、无线通信、微机电系统等学科领域,是21世纪以来迅速发展起来的一种新型网络。WSN在监测区域内部署大量传感器节点,先通过传感器协作地实时监测、感知和采集目标信息;然后将采集到的数据以无线通信技术方式传递;最后通过传感器间的多跳、自组织方式传送到用户终端。近年来,随着对无线传感器的研究以及不断完善WSN的网络协议,WSN已具有低功耗、低成本、分布式和自组织等优点,使其被广泛应用于军事系统、国家安全、环境监测、目标跟踪、智能家居、医疗保健和空间探索等领域。定位技术作为WSN中的关键技术之一,也是目前研究的热点。目前,WSN节点定位主要应用于静止状态下的目标定位,而对移动WSN定位研究比较少。如果在静态的WSN中部署若干个具有高速移动能力的节点,那么既增加了WSN定位的应用范围和应变能力,也满足了当今时代人们对位置信息的需求。因此,无论是在普通环境中的移动节点定位还是在高速环境中的节点定位都具有重要的理论意义和应用价值。本文对目前已有的移动WSN节点定位算法进行深入研究,并分析算法中存在的问题。针对定位区域的范围、定位精度和能耗等问题,结合各类移动定位算法具有的特点,提出一种在高速环境下的移动节点定位算法,利用不会变化的采样范围和航位推算算法具有短时间内定位精度高和独立性强的特点,进而得到精确的定位信息。本论文的主要研究成果如下:(1)针对蒙特卡罗定位算法采样范围过大这一现象,将线性部署与基于正三角形的节点部署结合,提出一种新的算法模型。利用锚节点感知范围可以重叠的特点,得到一个精确的重叠区域,并用其取代了传统蒙特卡罗定位算法以速度为半径确定的采样范围。(2)针对蒙特卡罗定位算法中粒子退化问题导致的重采样以及航位推算算法误差积累的问题,本文提出来一种改进航位推算算法,将重叠区域内采集到的锚节点信息作为修正定位误差的参照点。在离散的时间段内,利用参照点修正定位误差,并且利用定位误差描绘出未知节点的运动轨迹。(3)考虑到线性部署单点失效问题以及高速环境下移动未知节点的定位精度,本文引入数据回传技术与异常数据剔除技术,并在算法模型中虚拟地构建一个数据处理中心。高速移动未知节点将采集到的信息直接回传到数据处理中心,解决了线性部署中数据单向传递问题。经过修正因子与阈值对比,剔除大于阈值的误差值,进而得到更精确的定位。(4)对基于WSN的高速移动节点定位算法所提出的解决方案进行仿真,并采用匀速运动、匀加速运动和匀减速运动以三组不同的速度、加速度、参照点数进行比较分析。最后,对比分析了在一维空间与二维空间下算法定位误差,仿真结果表明在一维空间下的定位误差要小于在二维空间中的定位,同样说明了该算法同样适用于在一维空间下对高速移动节点的定位。
[Abstract]:Wireless Sensor Network (WSN) integrated the field of embedded computing technology, sensor technology, wireless communication and micro-electro-mechanical system. It is a new kind of network that has developed rapidly since the 21st century. The WSN deploys a large number of sensor nodes in the monitoring area, monitors, senses and collects the target information in real time through the sensors, and then transmits the acquired data in a wireless communication technology mode; and finally, through the multi-hop among the sensors, the self-organizing mode is transmitted to the user terminal. In recent years, with the research of the wireless sensor and the continuous improvement of the network protocol of the WSN, the WSN has the advantages of low power consumption, low cost, distributed and self-organization and the like, and is widely applied to the military system, the national security, the environment monitoring, the target tracking and the intelligent home, Healthcare and space exploration. As one of the key technologies in the WSN, the location technology is also the hot spot of the research. At present, the location of the WSN node is mainly applied to the target location in the stationary state, and the positioning of the mobile WSN is less. If a number of nodes with high-speed movement capability are deployed in the static WSN, the application range and the adaptability of the WSN positioning are increased, and the requirements of people on the position information in the current era are also met. Therefore, both the positioning of the mobile node in the normal environment or the positioning of the nodes in the high-speed environment have important theoretical and application values. In this paper, the existing mobile WSN node positioning algorithm is deeply studied and the problems existing in the algorithm are analyzed. Aiming at the problems of the range of the positioning region, the positioning accuracy and the energy consumption and the like, the mobile node positioning algorithm in a high-speed environment is proposed in combination with the characteristics of various mobile positioning algorithms, The method has the characteristics of high positioning accuracy and strong independence in a short time by utilizing the sampling range and the dead reckoning algorithm which do not change, and further obtains accurate positioning information. The main research results of this paper are as follows: (1) For the phenomenon that the sampling range of the Monte-Carlo positioning algorithm is too large, the linear deployment is combined with the node deployment based on the regular triangle, and a new algorithm model is proposed. A precise overlap region is obtained by using the characteristic of the sensing range of the anchor node, and the sampling range determined by the speed as the radius is replaced by the traditional Monte Carlo positioning algorithm. (2) In order to solve the problem of the re-sampling and the error accumulation of the dead reckoning algorithm caused by the problem of particle degradation in the Monte-Carlo location algorithm, this paper proposes an improved dead reckoning algorithm, which uses the anchor node information collected in the overlapping area as the reference point for correcting the positioning error. In the discrete time period, the positioning error is corrected by the reference point, and the motion trail of the unknown node is depicted by using the positioning error. (3) Considering the single point failure of the linear deployment and the positioning accuracy of the moving unknown node in the high-speed environment, this paper introduces the data return technique and the abnormal data rejection technique, and constructs a data processing center in the algorithm model. The high-speed moving unknown node transfers the collected information directly to the data processing center, and solves the problem of one-way transfer of data in the linear deployment. After the correction factor is compared with the threshold value, the error value larger than the threshold is eliminated, and the more accurate positioning is obtained. (4) The solution of the high-speed moving node positioning algorithm based on the WSN is simulated, and the uniform speed motion, the uniform acceleration motion and the uniform deceleration motion are compared and analyzed in three groups of different speed, acceleration and reference points. Finally, the positioning error of the algorithm in one-dimensional space and two-dimensional space is compared and analyzed, and the simulation results show that the positioning error in one-dimensional space is less than that in the two-dimensional space, and the same applies to the positioning of the high-speed moving node under the one-dimensional space.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5;TP212.9

【参考文献】