基于超宽带雷达的生命体征信息检测研究

发布时间：2019-03-23 15:47

【摘要】：城市化进程的快速发展带来了医疗、社会治安等方面的问题,如何很好的解决这些问题成为众多学者讨论和研究的热点。其中,非接触式体征检测作为一种重要的方法为解决此类问题提供了一个研究方向。而UWB雷达因其穿透能力强、抗干扰以及分辨率高等优点,使其在非接触式体征检测领域发挥着重要的作用。因此,研究基于UWB雷达的体征信号检测在非接触性体征信号检测领域具有重要的意义。本文从信号处理方面,以模式分类为指导思想,根据体征信号中各个成分的信号模式特征分别在直达波的消除、多目标的识别、目标信号的分离重组等方面进行体征信号的检测研究。在分析体征信息检测的主要技术路线和典型算法的基础上,提出了本文的技术总路线。本文的研究主要涉及以下内容:首先,综述了几种典型的非接触性检测方法及基于雷达的体征检测算法,并比较了各自的优缺点,在对UWB雷达体制的分析的基础上对UWB雷达在体征检测中的特点进行了分析,建立了生命体征信号模型。同时,总结了两种算法质量评价标准,分别从整体和局部的角度对算法进行质量评价。其次,对直达波进行分析并建立了直达波数学模型,提出了结合K-L变换的FastICA的直达波抑制算法。通过对静止的单人目标和运动的单人目标的实验研究,验证了该方法在直达波抑制方面的有效性,并通过与ICA、K-L的迭代次数、运行时间以及载干比、信噪比等客观性能的对比分析,证明了该算法在迭代速度和计算复杂度方面的优越性。再次,对体征信号中的呼吸、心跳成分,在对传统信号提取方法的分析研究的基础上,针对传统EMD方法端点飞跃和传统EEMD重构信号质量差等不足研究基于EEMD与BP优化策略的体征信号提取算法,并对单人体征信号的提取进行了实验研究,实验证明该算法不仅克服了EMD端点延拓问题,避免了端点飞跃现象,而且提高了重构信号的信噪比,验证了该算法适合于体征信号中具体信号成分的提取。最后,分析了多目标人体信号特征并构建了双人信号模型,并在此基础上对MUSIC方法作了理论推导,研究本文MUSIC算法的理论模型和算法实现步骤。分别对单人和双人静止人体目标信号进行实验研究,实验结果表明,该方法能有效地识别出人体目标信号数目,尤其是在多人体目标信号识别方面,具有分辨率高,抗干扰能力强等优点。
[Abstract]:The rapid development of urbanization has brought the problems of medical treatment, social security and so on. How to solve these problems well has become a hot topic discussed and studied by many scholars. Non-contact physical sign detection as an important method provides a research direction for solving this kind of problem. Because of its strong penetrating ability, anti-jamming and high resolution, UWB radar plays an important role in the field of non-contact physical sign detection. Therefore, it is of great significance to study the detection of physical sign signal based on UWB radar in the field of non-contact physical sign signal detection. From the aspect of signal processing, according to the pattern classification as the guiding ideology, according to the signal pattern characteristics of each component of the sign signal, the direct wave elimination and multi-target recognition, respectively, are carried out in this paper. The separation and recombination of the target signal is used to detect the physical sign signal. On the basis of analyzing the main technical route and typical algorithm of physical sign information detection, the general technical route of this paper is put forward. The main contents of this paper are as follows: firstly, several typical non-contact detection methods and radar-based physical sign detection algorithms are summarized, and their advantages and disadvantages are compared. Based on the analysis of UWB radar system, the characteristics of UWB radar in physical sign detection are analyzed, and the vital sign signal model is established. At the same time, the quality evaluation criteria of two algorithms are summarized, and the quality evaluation of the algorithm is carried out from the whole and local angles respectively. Secondly, the direct wave is analyzed and the mathematical model of the direct wave is established, and the direct wave suppression algorithm of FastICA based on K _ (L) transform is proposed. The effectiveness of this method in direct wave suppression is verified by experiments on static single-person targets and moving single-person targets, and the iterations with ICA,K-L, running time and carrier-to-interference ratio are also verified. The advantages of the algorithm in iteration speed and computational complexity are proved by comparing and analyzing the objective performance such as signal-to-noise ratio (SNR). Thirdly, on the basis of the analysis and research of the traditional signal extraction methods, the respiratory and heartbeat components in the sign signal are analyzed. Aiming at the deficiency of traditional EMD method such as endpoint leap and bad quality of traditional EEMD reconstruction signal, the physical sign signal extraction algorithm based on EEMD and BP optimization strategy is studied, and the extraction of single sign signal is studied experimentally. Experimental results show that the algorithm not only overcomes the problem of EMD endpoint continuation and avoids the phenomenon of endpoint leap, but also improves the signal-to-noise ratio of reconstructed signals. It is verified that the algorithm is suitable for the extraction of concrete signal components from physical signs. Finally, the signal characteristics of multi-target human body are analyzed and a double signal model is constructed. On this basis, the MUSIC method is derived theoretically, and the theoretical model and the algorithm realization steps of the MUSIC algorithm in this paper are studied. The experimental results show that the method can effectively identify the number of human body target signals, especially in the multi-body target signal recognition, and has high resolution, and the experimental results show that the proposed method can effectively identify the number of human body target signals, especially in the multi-human body target signal recognition. Strong anti-jamming ability and other advantages.
【学位授予单位】：山东建筑大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN957.51

【相似文献】