Study on Human Pulse Wave Measurement Based on Graphene-Based Flexible Sensors

针对人体脉搏波的测量问题，文章提出了基于石墨烯基柔性应变传感器的技术方案，并对数字滤波算法展开了研究。以惠斯通电桥作为传感器的基本测量电路，通过可调放大电路、带通滤波电路、直流偏置电路和模数转换电路测得脉搏波。为了进一步提取有用信息，采用了变分模态分解与平滑滤波相结合的数字滤波算法。利用所述方案对实际人体脉搏波进行了实验测试，结果与标准值最大偏差为1.63%，表明所提出的方案能够准确实现人体脉搏波的测量，调理后的脉搏波信号能清晰地观测到脉动、潮波和重搏波等特征，证明了基于石墨烯基柔性传感器的脉搏波测量方案的可行性。提取调理后的脉搏波信号特征，并使用基于改进粒子群算法的BP神经网络实现预测血压信号，进而预测得到人体血压的收缩压和舒张压，结果与实测值最大误差为4.09%，表明所提方案能准确测量人体血压。To measure human pulse waves precisely, a technical solution based on graphene-based flexible strain sensors is proposed, and a digital filtering algorithm is investigated. A Wheatstone bridge is used as the sensor’s basic measurement circuit, and the pulse wave is measured by an adjustable amplifier circuit, a bandpass filter circuit, a DC bias circuit, and an analog-to-digital converter circuit. To further extract useful information, a digital filtering algorithm combining variational mode decomposition and smoothing filter was designed. The actual human pulse wave measurement experiments using the described scheme were carried out. The maximum deviation between the result and the standard value is 1.63%, and the results show that the proposed scheme can accurately realize the measurement of the human pulse wave, and the features of pulsation, tidal wave, and repetition beat wave can be clearly observed in the conditioned pulse wave signal, which proves the feasibility of the proposed method. Extracting the features of the conditioned pulse wave signals and using a BP neural network based on an improved particle swarm algorithm to realize the blood pressure signals, and then obtaining the systolic and diastolic blood pressure of the human blood pressure, the result has a maximum error of 4.09% from the measured value, which shows that the proposed scheme can accurately measure the blood pressure of the human body.