隔墙人体运动检测的信号处理方法和系统设计
发布时间:2018-02-16 04:39
本文关键词: 隔墙 人体运动检测 信号处理 软件无线电 出处:《浙江大学》2017年硕士论文 论文类型:学位论文
【摘要】:近十年来的隔墙人体运动检测技术主要使用超宽带信号,但往往需要用GHz的带宽、较高的功率才能分辨出墙后的物体,且一般雷达系统体积较大。如果能够不使用超宽带信号,而采用基于工业、科学和医药(Industrial Scientific Medical,ISM)频段信号来实现一种小带宽、小功率、小体积、低成本的隔墙人体运动检测系统,将对社会公共安全以及人民生活的方方面面产生极大的影响。本文对隔墙人体运动检测技术进行了深入探究,设计实现了基于ISM频段信号的隔墙人体运动检测系统。本文主要工作包括:1.设计了隔墙人体运动检测系统。首先研究墙体等强背景噪声干扰下的隔墙人体运动检测机理;然后采用IEEE 802.11n信号进行通信,并利用预编码来消除墙体等静态物体反射回来的强干扰信号,从而能够提取出墙后微弱的人体运动回波信号;最后使用软件无线电设备及平台实现了信号收发以及背景噪声干扰抑制,并对干扰抑制效果进行分析,得出其抑制效果有一个上限。2.研究隔墙人体运动检测的离线信号处理方法。通过对接收信号进行时域、频域和联合时频域信号处理与分析比较,可以提取出用于隔墙检测使用的特征属性,进而提出了基于阈值的隔墙人体运动检测方法。为了适用于不同的检测环境,引入机器学习的思想,提出了基于聚类和分类的隔墙人体运动自适应检测方法。3.研究隔墙人体运动朝向检测的离线信号处理方法。从接收信号中提取出信道状态信息,进而获取多普勒频移以判断运动人体相对接收机的运动方向;且通过频偏可以大致估算出人体的运动速度。另外,通过使用动态时间规整算法能够检测区分出沿墙水平和垂直运动两种运动朝向。4.研究隔墙人体实时检测的信号处理方法,并设计结果展示界面。首先通过对离线方法进行修改,并辅助轻量级检测方法,能够给出更短时间内的运动模式信息;然后设计了结果展示界面及离线分析工具;最后对检测准确性进行了分析讨论。
[Abstract]:In recent ten years, the human motion detection technology of partition wall mainly uses UWB signal, but it often needs the bandwidth of GHz and the higher power to distinguish the object behind the wall, and the general radar system is larger. If UWB signal can not be used, A small bandwidth, small power, small volume and low cost partition wall human motion detection system is realized by using industrial, scientific and pharmaceutical industrial Scientific frequency band signals. Will have a great impact on social public safety and all aspects of people's life. A human motion detection system based on ISM frequency band signal is designed and implemented. The main work of this paper includes: 1. A human motion detection system is designed. Firstly, the mechanism of human motion detection under strong background noise is studied. Then the IEEE 802.11n signal is used to communicate, and the strong interference signal reflected from the static object such as the wall is eliminated by precoding, which can extract the weak echo signal of human body behind the wall. Finally, the software radio equipment and platform are used to realize the interference suppression of signal transceiver and background noise, and the effect of interference suppression is analyzed. It is concluded that there is an upper limit. 2. The off-line signal processing method for human motion detection of partition wall is studied. The signal processing in time domain, frequency domain and joint time and frequency domain is compared with the received signal, and compared with the received signal in time domain, frequency domain and joint time and frequency domain. The feature attributes used in partition wall detection can be extracted, and then a threshold based human motion detection method is proposed. In order to be suitable for different detection environments, the idea of machine learning is introduced. In this paper, an adaptive detection method of human motion based on clustering and classification is proposed. 3. The off-line signal processing method for detecting human motion orientation of partition wall is studied. The channel state information is extracted from the received signal. Then the Doppler frequency shift is obtained to judge the moving direction of the moving human body relative to the receiver, and the motion velocity of the human body can be roughly estimated by the frequency offset. By using dynamic time warping algorithm, we can detect and distinguish two kinds of moving directions. 4. The signal processing method of real time detection of human body with partition wall is studied, and the result display interface is designed. Firstly, the off-line method is modified. With the help of lightweight detection method, the motion mode information can be given in a shorter time. Then the result display interface and off-line analysis tool are designed. Finally, the accuracy of the detection is analyzed and discussed.
【学位授予单位】:浙江大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:TN911.7
【参考文献】
相关硕士学位论文 前4条
1 王海静;基于RSSI的无线传感器网络的人体探测[D];山东大学;2014年
2 张余;混沌频率步进穿墙雷达的设计与实现[D];南京航空航天大学;2013年
3 甄东;基于ZigBee平台的室内人员探测问题研究[D];山东大学;2013年
4 王涵宁;步进频连续波雷达穿墙成像技术研究[D];国防科学技术大学;2011年
,本文编号:1514679
本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/1514679.html