智能传感器网络中地面目标磁敏信号识别研究

发布时间：2018-04-03 01:23

本文选题：无线传感网　切入点：磁传感器　出处：《上海交通大学》2014年硕士论文

【摘要】：进入新千年来,随着各学科技术转化为产业化应用及传统互联网相关技术应用的日趋完善和成熟,各项技术集大成者的无线传感网(WSN)顺利地走上了历史舞台,并成为国内外各界追逐和研究的又一热点。WSN(Wireless Sensor Networks)最早源于军事领域的应用,最为典型的是可以对地面移动和低空飞行等目标进行探测和分类的无人地面传感器侦察网络。所以,传感节点对于目标信号的分类、跟踪、定位等处理能力一直是国际上广泛研究的热点课题之一。尽管如此,传感网对于目标的识别、分类和跟踪等的处理仍然是一项十分复杂的工作,因为在很大程度上地面传感网系统会受到外界环境的严重影响。地面传感器能在其覆盖范围之内尽最大可能地采集和反馈目标、环境等信息,并自主地完成相应的算法处理,将有助提高目标处理和识别的精准度与及时度,对后面的势态评估和任务决策有着重要的意义。与发达国家相比,我国的地面磁目标识别装备及系统相对比较落后,从九十年代初期开始,在自然基金的资助和申请研究课题的推动下,多项磁地面探测系统预研才能得到顺利开展,并取得相应的学术成果和技术的积累。这里明确了课题内容在系统框架拓扑结构中的位置,制定本课题研究的传感器的特性要求和实现的功能;深入调研了国内外现有目标识别磁信号传感器;比较国内现有各类应用在目标识别上的信号处理算法的优缺点。针对地面运动物体产生的磁异信号的特征信号提取和判决算法以及传感器设计进行了比较深入且系统的讨论。首先从地面运动物体会产生磁信号的本质入手,分析了地面各种类型目标产生的磁信号特征和原理;并建立地面目标沿传感器运动的模型,研究特征提取方法。随后主要在软件和硬件方面,设计一个针对地面目标磁信号识别的传感器,基于该传感器工作原理,设计一个用于目标磁异信号特征的目标检测和目标运动方向识别算法,并且通过实验和外场测试验证算法的可行性和系统的实际检测效果。本课题解决的关键问题是如何使检测算法对各类地面目标有较高的检测率和较低的误报率;比较于现有的算法和现有的地面目标传感器,该智能地面目标识别传感器明显地提升现有传感器技术性能,使其具有识别率高、稳定性高等优点,能更好地适合战场野外监控等需要大面积传感器应用的场合。
[Abstract]:Since entering the new millennium, with the increasingly perfect and mature application of various disciplines and technologies into industrial applications and traditional Internet-related technologies, WSNs, the wireless sensor networks (WSNs), have successfully stepped onto the historical stage.WSNN Wireless Sensor Networks (WSNN Wireless Sensor Networks) originated from the application of military field, and the most typical is the unmanned ground sensor reconnaissance network, which can detect and classify targets such as ground moving and low altitude flight.Therefore, the processing ability of sensor nodes for target signal classification, tracking, positioning has been one of the hot topics in international research.However, the processing of target recognition, classification and tracking in sensor networks is still a very complex task, because to a large extent, the ground sensor network systems will be seriously affected by the external environment.The ground sensor can collect and feedback the target, environment and other information as far as possible within its coverage area, and independently complete the corresponding algorithm processing, which will help to improve the accuracy and timeliness of target processing and recognition.It is of great significance to the later situation assessment and task decision.Compared with the developed countries, the equipment and system of ground magnetic target recognition in China is relatively backward. Since the early 1990s, it has been promoted by the funding of the Natural Resources Fund and the application and research projects.A number of magnetic ground detection systems can be developed successfully, and the corresponding academic achievements and technical accumulations are obtained.This paper clarifies the position of the subject content in the topological structure of the system framework, formulates the characteristic requirements and the functions of the sensors studied in this paper, deeply investigates the existing target recognition magnetic signal sensors at home and abroad.Compare the advantages and disadvantages of various signal processing algorithms used in target recognition in China.In this paper, the characteristic signal extraction and decision algorithm of the magnetic signal generated by the ground moving object and the design of the sensor are discussed in depth and systematically.Starting with the nature of magnetic signals produced by moving objects on the ground, this paper analyzes the characteristics and principles of the magnetic signals generated by various types of targets on the ground, and establishes a model of the ground objects moving along the sensor, and studies the method of feature extraction.Then, in the aspect of software and hardware, a sensor is designed to recognize the magnetic signal of the ground target. Based on the principle of the sensor, an algorithm for target detection and target direction recognition is designed.The feasibility of the algorithm and the actual detection effect of the system are verified by experiments and field tests.The key problem of this paper is how to make the detection algorithm have higher detection rate and lower false alarm rate for all kinds of ground targets, compared with the existing algorithms and existing ground target sensors,The intelligent ground target recognition sensor can obviously improve the technical performance of the existing sensors and make it have the advantages of high recognition rate and high stability. It can be better suited to the field monitoring and control of battlefield and other situations where large area sensors are needed.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP212.6;TN911.7

【相似文献】