Abstract: Pulse diagnostic acquisition devices mainly rely on hardware collecting data at high sampling rate to calculate physiological information such as pulse rate and pulse interval accurately, which needs complex equipment and requires a lot of storage space. Aiming at the problem, the paper designs a simple pulse diagnosis instrument to collect pulse data and recognizes pulse wave considering pulse characteristics and adaptive threshold to optimize the main pulse wave data based on principles of curve fitting. This method solves the over-saturation and data flutter phenomenon in the process of collecting data by software filtering, and can achieve accuracy sampling at 6 kHz by sampling at 60 Hz which could recognize pulse wave correctly over 99.93%. The Matlab simulation results show that, compared to the classical differential threshold method, the method proposed in this paper makes less errors in calculating the pulse interval and pulse rate, and increases accuracy of pulse?peak position and storage space utilization rate, which has very high engineering practicability.

Key words: pulse main wave fitting, data analysis, least squares, peak estimate

摘要： 目前脉搏信号采集系统以高采样率获取脉搏数据进而实现精确计算脉搏间期和脉率等生理信息，这对硬件设备的要求高，并且需要大量的存储空间。针对该问题，设计一款脉搏采集装置获取脉搏数据，综合脉搏波形特征和自适应阈值进行脉搏波识别，结合曲线拟合原理提出一种基于软件层面对脉搏主波数据进行优化处理的方法。该方法通过软件滤波达到平滑脉搏波形的效果，解决了数据采集过程中产生的过饱和现象和脉搏波不平滑问题，同时能够以60 Hz的采样率达到6 kHz采样率的精度，脉搏识别率达到99.93%以上。通过Matlab仿真实验表明，相比于经典的差分阈值法，该方法实现的脉搏间隔和脉率误差降低，精确度明显提升，节省了存储空间，具有很高的工程实用性。

关键词: 脉搏主波拟合, 数据分析, 最小二乘, 峰值估计