基于无线通信的多点脉搏信号检测方法研究

发布时间：2018-05-16 11:18

本文选题：PVDF + ZigBee无线技术　；参考：《华南理工大学》2012年硕士论文

【摘要】：人体的脉搏波动中蕴藏着丰富的生理病理信息，通过对提取的脉搏信号进行分析，可以将潜在的心血管疾病尽早的诊断出来，为心血管疾病的预防和治疗争取宝贵的时间。目前人们对脉搏信息检测和分析方面的研究做了很多工作，也取得了很多优秀的成果。基于无线通信的多点脉搏信号检测是一种新颖而有效的检测方法，它将PVDF压电薄膜传感器、Zigbee无线通信、USB接口和信号处理等技术相结合，使得检测过程更加方便灵活，检测结果更为准确丰富。使用五点式PVDF压电脉搏传感器来检测人体桡动脉处五点的脉搏信息，消除了单点传感器测量带来的偶然误差，而且获取的脉搏信息更为丰富。设计PVDF压电脉搏传感器为台型，将五个圆形的PVDF压电薄膜嵌入五个排列规则且互相独立的槽中，制作信号调理电路来处理PVDF压电薄膜形变产生的脉搏电荷信号，获取所需的脉搏电压信号。在传感器背侧安装电池和天线，保证信号采集过程中的电源供应与信号发射。采用Zigbee无线通信技术来传输脉搏信息，达到了移动性、实时性、可靠性、功耗低、成本低和辐射小的效果。使用USB接口连接PC机，实现了高速传输和方便快捷的目的。Zigbee单片机型号选用CC2430，USB接口芯片采用EZ-USB FX2，工作模式为SLAVE FIFO。数字脉搏信号从Zigbee网络节点传输到主机的过程是：Zigbee网络节点将采集的脉搏信号通过组网传给Zigbee协调器，协调器单片机CC2430作为外部主控制器控制EZ-USB FX2的工作，将脉搏数据通过USB接口传到PC机。SLAVE FIFO的工作模式降低了EZ-USB FX2接口芯片工作的复杂性，同时也提高了传输效率。获取的脉搏信息需要进行数字滤波处理。文中使用了三种滤波方法，即采用单片机程序对脉搏信息进行滑动算术平均值滤波、小波变换阈值滤波和建立在零点极点抵消基础上的简单整系数滤波，并对三种滤波效果进行了比较。接下来对获取的脉搏信息分别进行了时域和频域的信号分析，获得了较好的处理效果。时域方面，采用了基于模极大值的时域特征点提取和基于经验模态分解的时域特征点提取方法；频域方面，应用了基于功率谱估计的特征提取和基于谱能比的特征提取方法。
[Abstract]:There are abundant physiological and pathological information in pulse fluctuation of human body. By analyzing the extracted pulse signals, potential cardiovascular diseases can be diagnosed as soon as possible, and valuable time can be gained for the prevention and treatment of cardiovascular diseases. At present, people have done a lot of research on pulse information detection and analysis, and achieved a lot of excellent results. Multi-point pulse signal detection based on wireless communication is a novel and effective detection method. It combines PVDF piezoelectric film sensor with PVDF interface and signal processing technology, which makes the detection process more convenient and flexible. The results are more accurate and abundant. The five-point PVDF piezoelectric pulse sensor is used to detect the five-point pulse information of the human radial artery, which eliminates the accidental error caused by the single-point sensor measurement, and the obtained pulse information is more abundant. The PVDF piezoelectric pulse sensor is designed. Five circular PVDF piezoelectric films are embedded in five regular and independent grooves. A signal conditioning circuit is made to process the pulse charge signal generated by the PVDF piezoelectric film deformation. Obtain the desired pulse voltage signal. Batteries and antennas are installed on the back of the sensor to ensure power supply and signal transmission during signal acquisition. Zigbee wireless communication technology is used to transmit pulse information, which achieves the effect of mobility, real-time, reliability, low power consumption, low cost and low radiation. Using USB interface to connect PC, the purpose of high speed transmission and convenience is realized. EZ-USB FX2 is used in CC2430U USB interface chip, and the mode of operation is SLAVE FIFO. The process of transmitting digital pulse signal from Zigbee network node to host computer is that the pulse signal collected by the network node is transmitted to the Zigbee coordinator through the network, and the single chip microcomputer CC2430 acts as the external master controller to control the work of EZ-USB FX2. The mode of transmitting pulse data through USB interface to SLAVE FIFO reduces the complexity of EZ-USB FX2 interface chip and improves the transmission efficiency. The obtained pulse information needs to be processed by digital filtering. In this paper, three kinds of filtering methods are used, that is, sliding arithmetic average filter for pulse information, wavelet transform threshold filter and simple integer coefficient filter based on zero pole cancellation. Three filtering effects are compared. Then the obtained pulse information is analyzed in time domain and frequency domain, and good processing effect is obtained. In time domain, time domain feature point extraction based on modulus maximum and time domain feature point extraction based on empirical mode decomposition are adopted. In frequency domain, power spectrum estimation based feature extraction and spectral energy ratio based feature extraction method are applied.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318

【参考文献】