高机动目标组网协同探测技术
发布时间：2018-10-20 07:20
【摘要】：基于自组织传感器网络的目标跟踪技术无论在军事还是民用上都有着很大的应用价值,其主要任务是在噪声空间中分布位置分散的传感器节点,各节点对目标信息进行分析、交换、汇总,协同实现对目标的跟踪监测。如何在资源有限的情况下,合理地部署和管理传感器节点,使得目标区域没有监测盲区,并且让传感器组网能够在恰当的时间点上对进入监测范围的每一个目标指派对应的传感器进行稳定有效的跟踪是自组织网络设计的主要问题。本文根据自组织网络协同探测功能执行顺序,对相关内容进行了研究:组网中的传感器需要有科学的部署算法对其进行调动,在一定约束条件下,将组网监测范围覆盖整个监测区域,优化节点拓扑结构,并根据节点分布状态分派合适的传感器对其进行跟踪监测,以达到任务性能最优;完成传感器部署任务后,需要在噪声环境中根据空间位置分散的移动传感器或者固定传感器获得的传感器数据,设计合适的算法估计分析目标的运动参数;节点间通信是自组织网络能量主要消耗源,需要设计合理的算法简化节点间通信机制,同时还要提高组网对运动方向突变的目标的跟踪能力。本文的工作主要从这三个方面进行展开,主要内容如下:一、提出了应用于自组织网络的优化部署算法。部署算法中,每个节点根据局部监测区域节点密度和能量剩余情况,决定组网执行对等结构模式或能量排序模式,自主选择节能移动路径,改善节点拓扑结构,提升监测范围覆盖率,同时使得系统运行的持续时间更长久。二、对广泛应用的伪线性运动参数估计算法进行了改进。本文基于伪线性偏差补偿估计算法,构建辅助变量加权矩阵,减小了原算法的误差协方差矩阵,同时确保算法结果的收敛性。三、研究了一种分布式的二元量化变分算法,通过使用变分算法压缩在节点之间交换的数据,减少数据量,降低组网通信能耗。同时,二元量化变分算法具有良好的独立参数特性来确保跟踪性能稳定性,能够对具有轨迹突变的目标进行有效跟踪。
[Abstract]:The target tracking technology based on self-organizing sensor network has great application value in both military and civil fields. Its main task is to distribute the sensor nodes scattered in the noise space, and each node analyzes the target information. Exchange, summarize, and cooperate to achieve target tracking and monitoring. How to reasonably deploy and manage sensor nodes with limited resources, so that there are no monitoring blind areas in the target area, And it is the main problem in the design of ad hoc network that the sensor network can track the sensor assigned to every target entering the monitoring range at the right time. In this paper, according to the execution order of the cooperative detection function of the ad hoc network, the related contents are studied: the sensors in the network need to be mobilized by scientific deployment algorithm, and under certain constraints, The network monitoring area covers the whole monitoring area, and the topology of the nodes is optimized, and the appropriate sensors are assigned according to the distributed state of the nodes to track and monitor them, in order to achieve the optimal performance of the task, and after the sensor deployment task is completed, It is necessary to design a suitable algorithm to estimate the motion parameters of the target in noise environment according to the sensor data obtained by the spacial distributed moving sensor or the fixed sensor, and the communication between nodes is the main energy consumption source of the ad hoc network. It is necessary to design a reasonable algorithm to simplify the communication mechanism between nodes and to improve the ability of the network to track the moving direction sudden change target. The main contents of this paper are as follows: 1. An optimal deployment algorithm for ad hoc networks is proposed. In the deployment algorithm, each node decides to implement the peer-to-peer structure mode or the energy ranking mode according to the local monitoring of the regional node density and energy surplus, and independently chooses the energy-saving mobile path to improve the node topology. Increase the coverage of the monitoring range and make the system run longer. Secondly, the widely used pseudolinear motion parameter estimation algorithm is improved. Based on the pseudo-linear offset compensation algorithm, the auxiliary variable weighting matrix is constructed to reduce the error covariance matrix of the original algorithm and to ensure the convergence of the algorithm results. Thirdly, a distributed binary quantization variational algorithm is studied. By using the variational algorithm to compress the data exchanged between nodes, the amount of data is reduced and the communication energy consumption is reduced. At the same time, the binary quantization variational algorithm has good independent parameter characteristics to ensure the stability of tracking performance, and can effectively track the target with trajectory mutation.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5;TP212.9
本文编号：2282447
[Abstract]:The target tracking technology based on self-organizing sensor network has great application value in both military and civil fields. Its main task is to distribute the sensor nodes scattered in the noise space, and each node analyzes the target information. Exchange, summarize, and cooperate to achieve target tracking and monitoring. How to reasonably deploy and manage sensor nodes with limited resources, so that there are no monitoring blind areas in the target area, And it is the main problem in the design of ad hoc network that the sensor network can track the sensor assigned to every target entering the monitoring range at the right time. In this paper, according to the execution order of the cooperative detection function of the ad hoc network, the related contents are studied: the sensors in the network need to be mobilized by scientific deployment algorithm, and under certain constraints, The network monitoring area covers the whole monitoring area, and the topology of the nodes is optimized, and the appropriate sensors are assigned according to the distributed state of the nodes to track and monitor them, in order to achieve the optimal performance of the task, and after the sensor deployment task is completed, It is necessary to design a suitable algorithm to estimate the motion parameters of the target in noise environment according to the sensor data obtained by the spacial distributed moving sensor or the fixed sensor, and the communication between nodes is the main energy consumption source of the ad hoc network. It is necessary to design a reasonable algorithm to simplify the communication mechanism between nodes and to improve the ability of the network to track the moving direction sudden change target. The main contents of this paper are as follows: 1. An optimal deployment algorithm for ad hoc networks is proposed. In the deployment algorithm, each node decides to implement the peer-to-peer structure mode or the energy ranking mode according to the local monitoring of the regional node density and energy surplus, and independently chooses the energy-saving mobile path to improve the node topology. Increase the coverage of the monitoring range and make the system run longer. Secondly, the widely used pseudolinear motion parameter estimation algorithm is improved. Based on the pseudo-linear offset compensation algorithm, the auxiliary variable weighting matrix is constructed to reduce the error covariance matrix of the original algorithm and to ensure the convergence of the algorithm results. Thirdly, a distributed binary quantization variational algorithm is studied. By using the variational algorithm to compress the data exchanged between nodes, the amount of data is reduced and the communication energy consumption is reduced. At the same time, the binary quantization variational algorithm has good independent parameter characteristics to ensure the stability of tracking performance, and can effectively track the target with trajectory mutation.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5;TP212.9
【参考文献】
相关期刊论文 前2条
1 周全明;吴延凤;;定位系统在智能交通系统中的应用[J];计算机与数字工程;2011年08期
2 寇嘉梁;梁晓诚;郭红宾;;微生物传感器网络的研究[J];电脑与电信;2007年06期
，本文编号：2282447
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2282447.html
