基于小波变换的血压信号特征提取方法研究

该文针对传统的人体血压信号特征提取方法的不足，提出了一种了基于小波变换的人体血压信号特征量提取方法。人体的血压信号是一种非平稳的低频信号，使用Marr小波对人体血压信号进行小波变换的时间一尺度分解，可以得到人体血压信号的模极大值曲线；通过选择合适的分解尺度，可以有效地滤除噪声，对人体血压数据进行消噪处理；进一步滤除奇异性极大值点，可以提取出动脉波动位置和动脉波动幅度信息。文中给出了算法的完整实现，并利用实际采集的人体血压数据验证了该算法，得到了波峰模极大值点和心脏跳动所发生的精确时刻，为人体血压信号的参数识别奠定了基础。

Blood Pressure Signals Characteristics Extracting Method Based on Wavelet Transform

This paper focuses on a new characteristics extracting method of blood pressure signals based on wavelet transform in order to deal with dynamic blood pressure signals. Blood pressure signals are time - varying and low frequency. Applying Mart wavelet transform to blood pressure signals can get the wave crest modulus maxima points of blood pressure signals. The position information and amplitude information of artery are extracted from dynamic blood pressure signals by selecting proper scale for denoising and filtering the singular modulus maxima. A complete arithmetic is presented, the accurate time of heartbeats from the actual data can be located precisely. It establishes the foundation for recognizing the dynamic blood pressure signals.