一种基于智能床压电传感器的心率特征提取算法

A Feature Extraction Algorithm for Heart Rate Based on a Smart Bed Piezoelectric Sensor

介绍一种心率算法,应用于基于压电传感器的智能床系统。压电传感器安装在智能床床架上,并透过床垫接收人体振动信号。由于床垫厚度、材质差异,人体睡姿及传感器安装位置等影响因素,传感器采集到的人体振动信号波形差异很大。介绍的该心率算法,第1部分:信号预处理算法;第2部分:传感器信号存在3种主要情况,即心率信号较强,心率、呼吸信号并存,心率信号较弱,预处理后,信号一方面采用直接求解BCG(Ballistocardiogram)信号周期计算心率;另一方面采用求二阶导数,经验模态分解,希尔伯特变换算法,剥离出心率信号后,通过自相关算法求解BCG信号周期计算心率;第3部分:标记,评价计算过程产生的数据,估计最终心率。该方法能测量人在平躺时静息状态下的心率值,是一种有效的心率参数计算方法。

To introduce a heart rate algorithm applied to the smart bed system based on the piezoelectric sensor.Piezoelectric sensor is installed on the bedstead of smart bed and receives body vibration signal through the mattress.Due to different thicknesses,material mattress,body sleeping position and sensor installation position and other influencing factors,the human vibration signal waveform collected by the sensor is very different.The heart rate algorithm,the first part,is a signal preprocessing algorithm.The second part,due to three main conditions of sensor signal,namely,strong heart rate signal,coexisting heart rate and respiratory signal,and weak heart rate signal.After preprocessing,on the one hand,the heart rate was calculated by solving the BCG(Ballistocardiogram)signal directly,on the other hand,using the second derivative,empirical modal decomposition,hilbert transformation algorithm,stripping the heart rate signal,solving the BCG signal cycle with an autocorrelation algorithm to calculate heart rate.In the third part,marking,evaluates the data generated by the e calculation process,and estimates the final heart rate.The method can measure the heart rate in resting state and is an effective method to calculate heart rate parameters.