【作者】 刘海盆；

【导师】 黄春琳；

【作者基本信息】 国防科学技术大学 ， 电子与通信工程， 2010， 硕士

【摘要】 主动电磁波生命信号的实时检测处理是利用生命体对电磁波产生的多普勒效应,提取电磁波回波中的生命信号进行人体状态识别。根据人体的呼吸、心跳等生理特征,从主动发射电磁波的目标反射回波中,可以判别出人体有/无、动/静以及数量等信息。主动电磁波的生命信号探测方法,克服了传统的探生设备容易受外界环境限制和干扰的缺点,抗干扰能力强,适用于地震、塌方、山体滑坡等灾害后对受困人员的及时营救,同时也可用于医疗看护、狱所监管及反恐作战等领域,具有重要的应用价值。生命信号属于低速慢变信号,所产生的多普勒频移小。在灾后杂乱环境中检测生命体,回波信号微弱极易淹没于强杂波背景。如何有效地检测和提取出所需的微弱目标信号(即生命特征信号),以及识别出人体有/无、动/静以及数量等状态信息是论文的研究重点。本文根据小波变换的多分辨率特性,选取了d4小波,在噪声背景下检测和提取生命信号,采用呼吸与体动信号能量比、谐波参数估计及维格纳分布等方法,通过设置合适的处理门限值,自动完成人体状态信息的识别。该处理算法的综合处理能力强,实用性好,能够完成人体有/无、动/静以及数量等状态信息的判别,可以为操作人员提供直观、准确的信息。基于所研究的生命信号检测处理算法,设计了生命信号的实时检测处理系统。系统的硬件系统构成是分别以AD转换芯片AD7707和DSP处理芯片TMS320C6711B为核心,完成信号的解调和实时采集、传输及处理。该硬件系统设计电路功能合理,运行状态稳定,处理能力较强。基于TMS320C6711B高性能浮点DSP芯片,实现了数据的高速传输、实时处理和显示,使系统不仅体积小巧,抗干扰能力和探测能力等较之以C5000系列为DSP处理芯片的处理系统,均有所提高,并且为软件程序的稳定运行搭建一个良好平台。基于TMS320C6711B的生命信号实时检测处理系统,在CCS集成开发环境下完成了DSP的控制与实时检测处理软件程序的设计,主要包括生命信号实时检测处理算法的DSP软件设计与实现,以及Bootloader、系统初始化、中断服务、串口通讯、FIR滤波、FFT等程序的设计。该DSP部分的软件程序设计合理,能高效、实时的对生命信号进行采集处理,较好地实现了整个系统的控制和数据处理功能,并且实时显示出人体状态和数量等详细信息以及时频域的波形图。更多还原

【Abstract】 The technology of real-time detecting and processing active electromagnetic life signal is mainly based on the principle of Doppler frequency shifting. When the Electromagnetic wave was reflected by the body, we can extract life relevant characteristics information from the echo after properly being processed. Since signal was modulated by human life activities, such as breathing, heartbeating, etc, some parameters of the signal were changed. From the echo signal, we can estimate whether life is alive, active and the number of lives. The tradional life-detection equipments are always limited due to the external complicated environment, however,the method of real-time detecting and processing active electromagnetic life signal have the better anti-interference capability. This method is worth greater application, It not only can rescue injured persons quickly after earthquakes, collapse, mountain avalanche, and other disasters, but also can be used to medical care, prison surveillance as well as anti-terrorism fields and so on.Life signals are low-velocity and slow-change target signals with very small Doppler frequency shifts and so weak that echo signals were easily submerged in the serious clutter and noise. The key in this paper is to detect and extract the weak target signal, and recognize whether life is alive,active and the number of lives availably.This thesis designs a series of methods to recognize life signal from the echo. Firstly, the d4 wavelet which is based on the muti- resolution characteristic of wavelet transform was proposed to detect and extract life signal in the noise background. Secondly, the energy ratio of breathing and movement, parameters estimating of wave resonance and WD(Wigner Distribution)were designed to distinguish life relevant characteristics information automatically. This method has high efficiency and good practicability, and can provide the intuitionistic and credible information for the operator.Based on the designed algorithm framework, the life signal real-time detecting and processing system was designed. The kernel of the hardware part of the system is AD7707 and TMS320C6711B. The system can realize the high-speed transmiting, real-time processing and displaying of the data. It is not only tiny and compact, but also have better anti-interference and detecting ability. Moreover, it can provide the well hardware plat for the stable software running.Based on the TMS320C6711B, the software part of the system is designed under the CCS development software, which contains design and realization of DSP software, as well as code design of Bootloader, system initializing, interrupt service, data communication, FIR filtering, FFT and so on. The software codes of DSP part were designed reasonably, which make the life signal collecting and processing efficiently. It realized system control and data processing, and real-time display of results, as well as time-frequency waveform figures.更多还原