利用小波变换进行心率变异信号分形维数的短时估算

提出了一种利用功率谱短时估算心率变异信号分形维数的新方法。该方法基于小波变换和滤波器组，算法的实现过程是；首先利用小波变换的正交分解分离出心率变异信号中的分形成分；其次用自回归模型对分离出的分形成分进行功率谱估计，并用最小二乘法对分形成分的双对数坐标功率谱图进行直线拟合，估计出斜率7，最后根据公式D=2-（γ—1）／2估算出心率变异信号的分形维数。为了验证本文提出算法的稳定性和可靠性，采用分数布朗运动仿真24例分形维数值为1．6的分形信号进行验证，结果表明本算法应用于心率变异信号分形维数的估笪切空可行

A Method to Estimate the Short-term Fractal Dimension of Heart rate Variability Based on Wavelet Transform

A new method of calculating fractal dimension of short-term heart rate variability signals is presented. The method is based on wavelet transform and filter banks. The implementation of the method is: First of all we pick-up the fractal component from HRV signals using wavelet transform. Next, we estimate the power spectrum distribution of fractal component using auto-regressive model, and we estimate parameter 7 using the least square method. Finally according to formula D = 2- (gamma-1)/2 estimate fractal dimension of HRV signal. To validate the stability and reliability of the proposed method, using fractional brown movement simulate 24 fractal signals that fractal value is 1.6 to validate, the result shows that the method has stability and reliability.