对于煤矿作业者而言，井下作业环境中普遍存在高温、粉尘和噪声等职业危害因素，同时长期高强度和超负荷的劳动，容易增加心血管系统异常的风险。心率作为人体的一项重要生理指标，可以评估心脏活动能力和疲劳程度。然而，目前矿用心率监测设备存在集成度低、功耗高、运动伪影导致精度不高等问题。因此，本文设计一款矿用腕式心率监测系统，减少心率提取中的噪声干扰，在煤矿环境下实现人体心率参数的实时准确监测预警。本文主要工作内容如下：

对光电容积脉搏波描记法（Photoplethysmography，PPG）采集信号时的各种噪声影响，提出一种抗运动干扰的心率提取算法。首先将PPG信号状态分为平静状态和运动状态。其次对平静状态下的PPG信号进行预处理去除噪声，在运动状态下的PPG信号受运动伪影干扰严重，采用三轴加速度信号作为参考信号，将单级自适应滤波器改为加速度级联的递归最小二乘（Recursive Least Squares，RLS）自适应滤波器，去除PPG信号中的大部分运动伪影，接着通过自适应噪声完备集合经验模态分解（Complete Ensemble Empirical Mode Decomposition with Adaptive Noise，CEEMDAN）算法进行PPG信号的分解和重构，去除信号中残余的运动伪影。最后将去噪后的PPG信号统一在频域中进行心率提取，得到最终心率值。在公开数据集上进行实验，结果表明所提出算法在保持一定心率提取精准度的同时能有效降低计算复杂度。

对矿用腕式心率监测系统设计，包括硬件设计和软件设计，并进行终端APP设计。在硬件设计上，逐一进行主控蓝牙模块、PPG信号采集模块、加速度信号采集模块、人机交互模块和本安电源模块的电路设计。为保障井下使用的安全性，本安电源模块需进行防爆设计，以恒流型为核心增加双重保护电路设计，实现防反接和瞬态能量抑制以及过流恒流电路保护功能；供电转换电路采用Buck和LDO两种形式实现不同电压之间的转换，支持系统各模块正常工作。在嵌入式软件设计上，完成对主控、PPG信号采集、加速度信号采集和蓝牙传输等驱动程序的编写，并在终端APP上显示处理后的心率数据与预警情况。

对系统进行电气性能测试和功能测试，仿真与实际电路测试表明矿用腕式心率监测系统的电气性能上具有防反接、软启动、过流恒流保护和短路保护功能，能安全应用在井下环境；系统分别采集了4种不同状态下10名被测者的心率数据，与心率胸带的监测数据进行对比，结果表明心率提取算法能够嵌入设备终端，心率置信区间的概率为96.3%，当心率异常时能够及时预警，支持48小时以上工作时长。以上测试结果证明了系统具有高精度、抗干扰能力强和稳定性强的特点，适用于煤矿井下人员心率监测，具有一定的研究意义和实用价值。

For coal mine operators, occupational hazards such as high temperature, dust and noise are prevalent in the underground operating environment, while long-term high-intensity and overloaded labor tends to increase the risk of cardiovascular system abnormalities. Heart rate, as an important physiological indicator of the human body, can assess cardiac activity and fatigue. However, the current mining heart rate monitoring devices have problems such as low integration, high power consumption, and poor accuracy due to motion artifacts. Therefore, in this paper, we design a mining wrist heart rate monitoring system to reduce the noise interference in heart rate extraction and realize real-time accurate monitoring and warning of human heart rate parameters in coal mine environment. The main work of this paper is as follows:

To the photoelectric volumetric pulse wave tracing method (PPG) to collect the signal of various noise effects, put forward a heart rate extraction algorithm of anti-motion interference. Firstly, the PPG signal state is categorized into calm state and motion state. Secondly, the PPG signal in the calm state is preprocessed to remove the noise, and the PPG signal in the motion state is seriously interfered by motion artifacts, so the three-axis acceleration signal is used as the reference signal, and the single-stage adaptive filter is changed to the acceleration cascade Recursive Least Squares (RLS) adaptive filter, which removes most motion artifacts in the PPG signal, and then the PPG signal is extracted by the self-propagation algorithm, and the PPG signal is extracted by the self-propagation algorithm. Most of the motion artifacts are removed from the PPG signal, and then the PPG signal is decomposed and reconstructed by the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithm to remove the residual motion artifacts from the signal. Finally, the denoised PPG signal is unified in the frequency domain for heart rate extraction to obtain the final heart rate value. Experiments are conducted on public datasets, and the results show that the proposed algorithm can effectively reduce the computational complexity while maintaining a certain degree of accuracy in heart rate extraction.

The design of the mining wrist heart rate monitoring system includes hardware design and software design, and terminal APP design. In the hardware design, the circuit design of the main control Bluetooth module, PPG signal acquisition module, acceleration signal acquisition module, human-computer interaction module and intrinsically safe power supply module are carried out one by one. In order to protect the safety of underground use, intrinsically safe power supply module needs to carry out explosion-proof design, with constant current type as the core to increase the double protection circuit design, to realize the anti-reverse connection and transient energy suppression, as well as overcurrent constant current circuit protection; power supply conversion circuit adopts the two forms of Buck and LDO to realize the conversion between different voltages, to support the normal operation of each module of the system. In terms of embedded software design, the writing of drivers for main control, PPG signal acquisition, acceleration signal acquisition and Bluetooth transmission is completed, and the processed heart rate data and warning situation are displayed on the terminal APP.

Electrical performance testing and functional testing of the system, simulation and actual circuit testing shows that the electrical performance of the mining wrist heart rate monitoring system has anti-reverse connection, soft start, overcurrent constant current protection and short circuit protection, and it can be safely applied in the underground environment; the system collected the heart rate data of 10 subjects in 4 different states, and compared it with the monitoring data of the heart rate chest strap, and the results show that the heart rate extraction algorithm is The results show that the heart rate extraction algorithm can be adapted to the embedded device terminals, the probability of heart rate confidence interval is 96.3%, and the system can provide timely warning when the heart rate is abnormal, and it supports more than 48 hours of working hours. The above test results prove that the system is characterized by high precision, strong anti-interference ability and stability, and is suitable for heart rate monitoring of underground personnel in coal mines, which has certain research significance and practical value.