具有驾驶员特异性隐形可靠的心率信息监测理论与技术研究

Road safety has become one of the most serious problems in the world. Safe driving assistant (SDA) is an effective technology to reduce road accidents. Unobtrusive and reliable driver’s heart rate information monitoring is the key issue of SDA. This project will focus on the following points: First, study of technique for unobtrusive sensing heart rate related vital signal in vehicle and establish a research database; Second, by using the research database, study signal processing theory and technology for such vital signals, which include two parts: (1) Information theory based signal quality assessment and (2) signal quality level based multi-level signal enhancement technique with adaptive parameter adjustment. The signal processing theory and technology will ensure the reliability of information analysis for driver monitoring. Finally, study of personalized heart rate information analysis method is proposed, in which combine entropy model and personalized information analysis framework for accurately individual heart rate event prediction. The project aims to propose a series of theory and technique for unobtrusive reliable driver-specific heart rate information monitoring. It will provide theoretical and technical supports for accurate heart rate pattern recognition, SDA, driver health monitoring in vehicle.

道路安全已成为全球面临最为严峻的问题之一。安全驾驶辅助是有效降低道路交通事故的技术手段。车载环境下隐形可靠的驾驶员心率信息监测是安全驾驶辅助的核心问题。本项目重点研究内容包括：首先，研究车载隐形化心率相关信号采集技术并建立研究数据库；然后，在上述数据基础上进一步开展车载心率相关信号处理理论和技术研究。包括基于信息理论的信号质量评估和基于信号质量分级基础上的多层次参数可调的信号增强方法，保证车载心率分析的可靠性；最后，开展车载个性化心率信息分析技术研究。结合熵测度模型和个性化信息分析框架，研究个体车载心率事件精准预测方法。通过本项目的研究，提出完整的具有驾驶员特异性隐形可靠的心率信息监测理论和方法，为准确的心率模式识别，安全驾驶辅助，车载环境下健康监测提供理论基础和技术支持。

隐形可靠心率信息监测技术是基于生理信号驾驶员异常状态识别的基础和瓶颈。项目从隐形心率测量的生理信号传感技术、提高信号可靠性和可用性的信号质量评估和处理技术以及个体特异性信号处理技术三个方面开展了研究。项目在可长时间使用的穿戴式心电采集装置，图像光电容积描记法脉率估计，体表电势测量与躯干-心脏电活动模型有限元仿真的传感优化方法，信号质量评估方法和基于深度学习的心拍分类模型等方面取得了进展。实现了动态条件下心电信号持续采集，隐形化非接触车载心率监测，心电导联和位置优化，动态信号可用性和可靠性判别，以及个体特异性的心拍异常检测等。项目研究成果丰富了个体特异性隐形可靠心率信息监测方法，能够为可靠心率监测场景应用提供理论基础和技术支持。