Tianjin University of Technology and Education, Institute of Robotics and Intelligent Equipment, Tianjin 300222, China.
Sensors (Basel). 2019 Jan 17;19(2):368. doi: 10.3390/s19020368.
An unconstrained monitoring method for a driver's heartbeat is investigated in this paper. Signal measurement was carried out by using pressure sensors array. Due to the inevitable changes of posture during driving, the monitoring place for heartbeat measurement needs to be adjusted accordingly. An experiment was conducted to attach a pressure sensors array to the backrest of a seat. On the basis of the extreme learning machine classification method, driving posture can be recognized by monitoring the distribution of pressure signals. Then, a band-pass filter in heart rate range is adapted to the pressure signals in the frequency domain. Furthermore, a peak point array of the processed pressure frequency spectrum is derived and has the same distribution as the pressure signals. Thus, the heartbeat signals can be extracted from pressure sensors. Then, the correlation coefficient analysis of heartbeat signals and electrocardio-signals is performed. The results show a high level of correlation. Finally, the effects of driving posture on heartbeat signal extraction are discussed to obtain a theoretical foundation for measuring point real-time adjustment.
本文研究了一种驾驶员心率的无约束监测方法。信号测量采用压力传感器阵列进行。由于在驾驶过程中不可避免地会改变姿势,因此需要相应地调整心率测量的监测位置。进行了一项实验,即将压力传感器阵列附接到座椅的靠背。基于极限学习机分类方法,可以通过监测压力信号的分布来识别驾驶姿势。然后,在频域中适应带通滤波器以获取心率范围内的压力信号。此外,还可以得出处理后的压力频谱的峰值点数组,并且该数组与压力信号具有相同的分布。因此,可以从压力传感器中提取心率信号。然后,对心率信号和心电图信号进行相关系数分析。结果表明相关性很高。最后,讨论了驾驶姿势对心率信号提取的影响,为测量点实时调整提供了理论基础。