煤炭作为我国的主要能源在开采过程中可能出现矿山生产安全事故，造成人员被困 于井下的情况，需立即开展救援行动。当在救援过程中面临存在障碍物遮挡的情况时， 超宽带（Ultra-wide Band）雷达成为障碍物后方被困人员运动信息探测的有效技术手段； 而探测到人体运动信息的关键是对 UWB 雷达回波的高效处理，并从雷达回波中提取出 有用信号，以实现人体位置信息的获取。为此，本文以构建 UWB 雷达回波模型、优化 信号处理算法、结合仿真模拟与现场实验处理分析雷达回波为思路，开展 UWB 雷达回 波人体运动信息有效特征提取研究，为 UWB 雷达在矿山救援中的应用提供参考价值。

首先，通过分析人体运动学规律与建立人体坐标系，推导出人体行走过程中脊柱、 肩部、膝盖、肘部、手掌和脚掌等 9 个主要部位的运动数学模型；并将其与雷达方程组 结合，在确定中心频率为 400 MHz、波形为 ricker 函数调制、噪声干扰为-20 dB 高斯信 号的基础上，构建出贴近实际矿山救援环境的 6 种加载人体运动信息的 UWB 雷达回波 模型，包含：不考虑行走速度影响的脊柱、膝盖、肘部和肩部运动 UWB 雷达回波模型， 考虑人体行走速度为 1 m/s 影响的靠近雷达、远离雷达的躯干运动 UWB 雷达回波模型。

其次，利用 K-L 相对熵选取 VMD（Variational Mode Decomposition）算法中固有模 态分量个数 K、二次惩罚因子a的最优值分别为 8、3000，完成对 VMD 算法改进，并将 其与 ICA（Independent Component Analysis）算法结合，形成改进型 VMD-ICA 联合算法。

再次，将构建的 6 种 UWB 雷达回波模型分别通过 ICA 算法、改进型 VMD 算法、 改进型 VMD-ICA 联合算法的处理，并结合各算法处理雷达回波后信号的信噪比分析， 得出：三种算法对 UWB 雷达回波的处理性能由低到高依次为 ICA 算法（1.65 dB）、改 进型 VMD 算法（4.22 dB）、联合算法（6.55 dB，相对于两种算法单独使用时信噪比的 平均增长率为 342.37%、55.11%）；同时联合算法对躯干运动的 UWB 雷达回波处理效 果高于肢体运动，且可以综合两种算法的优势，实现对 UWB 雷达回波信号的高效处理。

最后，通过设计 UWB 雷达生命信息探测系统装置与搭建模拟矿山救援场景，开展 有无障碍物场景下，人体在距离雷达 2.5-3 m 范围内走动的 UWB 雷达穿透式探测现场 实验，从原始图谱特征中获得反应人体运动的波动持续时间为 2.70 ns；穿透不同厚度和 介质种类条件下，开展人体在不同范围内走动的现场探测实验，并将各 UWB 雷达回波 经联合算法处理后，进一步计算出人体位置信息，得到最大探测误差保持在 0.5 m 以内。

本文证实了改进型 VMD-ICA 联合算法在矿山救援场景中处理 UWB 雷达回波的性 能，为 UWB 雷达在矿山救援领域中的应用提供了重要的借鉴价值。

Coal serves as China’s primary energy source; however, safety accidents may occur during the mining process, causing personnel to become trapped underground and necessitating immediate rescue operations. When obstacles impede rescue efforts, ultra-wideband (UWB) radar proves to be an effective technical means of detecting the movement information of those trapped behind such obstacles. The key to detecting human motion information lies in efficiently processing UWB radar echoes and extracting useful signals from these echoes to obtain the location of individuals. Therefore, following the approach of constructing a UWB radar echo model, optimizing signal processing algorithms, and analyzing the radar echoes through both simulation and on-site experiments, this paper focuses on effectively extracting human motion features from UWB radar echoes. The findings aim to provide valuable insights for the application of UWB radar in mine rescue scenarios.

Next, by analyzing human kinematics and establishing a human coordinate system, this study derives the mathematical functions and models for nine major body parts—including the spine, shoulders, knees, elbows, palms, and soles—during walking. These models are then combined with a system of radar equations under conditions where the center frequency is set to 400 MHz, the waveform is modulated using a Ricker function, and the noise interference is - 20 dB Gaussian noise. On this basis, six types of UWB radar echo models that incorporate human motion information are constructed to reflect realistic mine rescue environments. These include: UWB radar echo models for the spine, knees, elbows, and shoulders without considering walking speed, as well as trunk-motion echo models that account for a 1 m/s walking speed—both approaching and receding from the radar.

Subsequently, the K-L relative entropy was used to determine the optimal values for both the number of intrinsic mode components (K) and the secondary penalty factor in the VMD algorithm, identified as 8 and 3000 respectively. This optimization completed the improvement of the VMD algorithm. In combination with the ICA algorithm, the improved VMD-ICA joint algorithm was thus established.

Thirdly, each of the six constructed UWB radar echo models was processed using the ICA algorithm, the improved VMD algorithm, and the improved VMD-ICA joint algorithm. By analyzing the signal-to-noise ratio (SNR) after each algorithm’s processing, it was found that the three algorithms’ performances in processing UWB radar echoes, from lowest to highest, were as follows: ICA (1.65 dB), improved VMD (4.22 dB), and the joint algorithm (6.55 dB), corresponding to SNR improvements of 342.37% and 55.11% over the ICA and improved VMD algorithms, respectively. Furthermore, the joint algorithm produced better processing results for trunk-motion UWB radar echoes than for limb-motion echoes, effectively combining the advantages of both algorithms to achieve efficient UWB radar echo signal processing.

Finally, by designing a UWB radar life-detection system and constructing a simulated mine rescue scenario, on-site experiments were conducted under both obstructed and unobstructed conditions with a person walking within 2.5–3 meters of the radar. From the raw spectrogram features, a fluctuation duration of 2.70 ns was identified as indicative of human movement. Under varying thicknesses and types of media, additional on-site detection experiments were performed with a person walking within different ranges. After the joint algorithm processed each UWB radar echo, the individual’s location information was further calculated, resulting in a maximum detection error kept within 0.5 meters.

This study confirms the performance of the improved VMD-ICA joint algorithm in processing UWB radar echoes within mine rescue scenarios, offering important reference value for the application of UWB radar in the field of mine rescue.

[1] 康红普,王国法,王双明,等.煤炭行业高质量发展研究[J].中国工程科学,2021,23(05): 130-138.

[2] 张俊文,杨虹霞.2005-2019 年我国煤矿重大及以上事故统计分析及安全生产对策研究[J].煤矿安全,2021,52(12):261-264.

[3] 宋志强,崔传波,邓存宝.巷道环境与产烟产热影响下的矿井火灾避灾路径选择[J/OL].矿业安全与环保.https://doi.org/10.19835/j.issn.1008-449 5.2023.05.02

[4] 郑学召,丁文,蔡国斌,等.面向钻孔救援的 UWB 雷达回波信息处理关键问题研究进展[J].煤矿安全,2023,54(10):219-225.

[5] 袁亮.深部采动响应与灾害防控研究进展[J].煤炭学报,2021,46(03):716-725.

[6] 王莉,胡一,王涛,等.基于改进相互作用矩阵-模糊认知图的矿井火灾风险评价[J].科学技术与工程,2023,23(13):5812-5818.

[7] 张盈盈,杨云,张景,等.2021—2023 年我国煤矿事故统计分析及致因研究[J].矿业研究与开发,2024,44(11):224-231.

[8] 程磊,孙洁. 2016—2022 年我国煤矿事故统计与规律分析[J].煤炭工程,2023,55(11): 125-129.

[9] 文虎,张铎,郑学召.矿山钻孔救援生命探测技术研究进展及趋势[J].煤矿安全,2017, 48(9):85-87.

[10] 郑学召,王虎,文虎,等.矿井钻孔救援通信技术的研究进展及趋势[J].工矿自动化,2017, 43(9):41-45.

[11] 郑学召,孙梓峪,郭军,等.矿山钻孔救援多源信息探测技术研究与应用[J].煤田地质与勘探,2022,50(11):94-102.

[12] 黄军利.国外煤矿应急救援技术装备综述[J].煤矿安全,2016,47(9):233-236.

[13] 李亮.平邑石膏矿坍塌事故救援成功后的几点思考[J].探矿工程(岩土钻掘工程), 2016, 43(10): 281-286.

[14] 王久平.“地心营救”既有收获也见不足-写在山东笏山矿难 11 名矿工获救之后[J]. 中国应急管理, 2021(2):10-13.

[15] 文虎,张铎,郑学召,等.基于 FDTD 的电磁波在煤中传播特性[J].煤炭学报,2017,42 (11):2959-2967.

[16] 黄泽华.基于 UWB 的室内定位系统研究[D].济南：山东大学,2021.

[17] 赵文生.综采工作面复杂环境下超宽带测距技术研究[J].煤炭科学技术,2019,47(12): 145-149.

[18] 文虎,刘盛铠,郑学召,等.基于钻孔救援的 UWB 雷达波传输衰减研究及展望[J].工矿自动化,2023,49(04):42-49.

[19] 郑学召,孙梓峪,张嬿妮,等.面向钻孔救援的超宽带雷达技术研究现状与方向[J].工矿自动化,2021,47(8):20-26

[20] 文虎,周博,郑学召,等.UWB 雷达在矿山钻孔救援中的应用研究[J].工矿自动化,2023,49(06):88-94.

[21] 吴文庆.矿井救援人员精确定位技术研究[D].西安：西安科技大学,2019.

[22] Pohorille A, Sokolowska J. Evaluating Biosignatures for Life Detection[J]. Astrobiology,

2020,20(10): 1236-1250.

[23] Y. S. Yoon and M. G. Amin, High-resolution through-the-wall radar imaging using Beam

space MUSIC, IEEE Transactions on Antennas and Propagation, vol.56, no.6, pp.1763-

1774, 2016.

[24] 杨东辉,甄杰,隋心.超宽带与激光雷达组合的室内定位方法[J].测绘科学,2019, 44(06):72-78.

[25] 陈新科,喻川,文智力.UWB 定位技术在煤矿井下的应用[J].煤炭科学技术,2018,46 (1):187-189.

[26] 刘世森.基于 UWB 的矿井人员精准定位技术[J].煤矿安全,2019,50(6):118-120.

[27] M.Kok, J.D.Hol, et al. “Indoor positioning using ultrawideband and inertial measurements,” IEEE Trans. eh. Technol., vol. 64, no.4, pp.1293-1303, Jan. 2015.

[28] P. K. Yoon, S. Zihajehzadeh, et al. “Robust biomechanical model-based 3-D indoor

localization and tracking method using UWB and IMU,” IEEE Sens. J., vol.17, no.4,pp.

1084-1096,Dec.2017.

[29] S. Zihajehzadeh and E. J. Park, “A novel biomechanical model-aided IMU/UWB fusion for

magnetometer-free lower body motion capture,” IEEE Trans. Syst., Man, Cy-bern.: Syst.,

vol. 47,no.6, pp. 927-938,Feb.2017.

[30] D. Feng, C. Wang, et al. “Kalman-filter-based integration of IMU and UWB for highaccuracy

indoor positioning and navigation,” IEEE Internet ThingsJ.,vol.7,no.4,pp.3133-

3146,Jan.2020

[31] 匡兵,陈凤冉,田春月,等.基于改进粒子滤波的超宽带与惯性测量单元融合定位算法[J].科学技术与工程,2020,20(30):12460-12466.

[32] 张斌飞,靳伍银.基于 UWB 和 IMU 紧组合的室内定位导航算法[J].电子测量技术, 2022,45(10):67-73.

[33] 古玉锋,杜雨洁,王育阳,等.基于神经网络与自调节卡尔曼滤波的 UWB 定位算法研究[J/OL].中国机械工程:1-11[2023-11-17].

[34] 杜浩国,孙志,杜浩标,等.震后基于无人机快速三维定位被困人员模型研究[J].震灾防御技术,2021,16(02):404-413.

[35] 柯新,黄建祥,薛质.基于超宽带雷达的三维成像技术研究[J].信息技术,2015(12):171 -174.

[36] 曹凌霄.穿墙雷达三维透视成像技术研究[D].成都：电子科技大学,2019.

[37] 郭继坤,赵清,徐峰.基于 SVM 的煤矿井下超宽带穿透成像算法研究[J].煤炭学报,2018,43(02):584-590.

[38] 郭继坤,赵清,陈司晗.基于相位补偿的矿井超宽带雷达压缩感知成像算法[J].煤炭科学技术,2020,48(1):211-218.

[39] 赵清.煤矿井下超宽带信号穿透塌方体成像算法研究[D].哈尔滨：黑龙江科技大学,2018.

[40] 赵尤信,姚海飞,李佳慧,等.超宽带雷达生命探测技术研究[J].工矿自动化,2023,49(9): 178-186.

[41] KUMAR A, ZHOU Yuqing, XIANG Jiawei. Optimization of VMD using kernel-based mutual information for the extraction of weak features to detect bearing defects[J]. Measurement,2021,168. DOI：10.1016/j.measurement.2020.108402.

[42] WANG Ran, XU Lei, LIU Fengkai. Bearing fault diagnosis based on improved VMD and DCNN[J].Journal of Vibroengineering,2020,22(5)：1055-1068.

[43] 刘长良,武英杰,甄成刚.基于变分模态分解和模糊 C 均值聚类的滚动轴承故障诊断[J].中国电机工程学报,2015,35(13):3358-3365.

[44] 康守强,胡明武,王玉静,等.基于特征迁移学习的变工况下滚动轴承故障诊断方法[J].中国电机工程学报,2019,39(3):764-772,955.

[45] 马增强,李亚超,刘政,等.基于变分模态分解和 Teager 能量算子的滚动轴承故障特征提取[J].振动与冲击,2016,35(13)：134-139.

[46] REN L Y, WANG H F, NAISHADHAM K, et al. Phase-based methods for heart rate detection using UWB impulse Doppler radar[J]. IEEE Transactions on Microwave Theory and Techniques,2016, 64(10):3319-3331.

[47] MOSTAFA M, CHAMAANI S, SACHS J. Applying singular value decomposition for clutter reduction in heartbeat estimation using M-sequence UWB Radar[C].2018 19th International Radar Symposium, Bonn, Germany,20-22 June,2018:1-10.

[48] LIANG X L, ZHANG H, FANG G Y, et al. An improved algorithm for through-wall target detection using ultra-wideband impulse radar[J].IEEE Access,2017,5:22101-22118.

[49] 郭继坤,王保生,郝维来,等.基于超宽带信号的矿井塌方体下生命特征的检测方法[J].黑龙江科技大学学报,2017,27(1):73-76.

[50] 马铖.面向低信噪杂比环境的随机信号雷达生命探测方法研究[D].太原：太原理工大学,2022.

[51] Khan F, Cho S. A detailed algorithm for vital sign monirtoring of a stationary/non -

stationary human through IR-UWB radar[J]. Sensors,2017,17,290.

[52] 刘永立,吴闻轩.井下超宽带人员定位的改进卡尔曼滤波算法[J].黑龙江科技大学学报,2021,31(4):500-505.

[53] 闫嘉明．基于非接触式生命体征探测雷达的人体目标识别理论与技术研究[D].南京：南京理工大学，2020.

[54] 任晶晶.超宽带信号穿墙传播特性建模与检测技术研究[D].合肥：中国科学技术大学,2012.

[55] 王保生.超宽带信号穿透塌方体衰减特性及生命特征检测方法研究[D].哈尔滨：黑龙江科技大学,2017.

[56] LI Jing，ZENG Zhaofa，SUN Jiguang，et al. Through-wall detection of human being's movement by UWB radar[J]. IEEE Geoscience and Remote Sensing Letters,2012,9(6)：

1-5.

[57] LI Jing，LIU Lanbo，ZENG Zhaofa，et al. Advanced signal processing for vital sign extraction with applications in UWB radar detection of rapped victims in complex environments[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2014,7：783-791.

[58] An Q, Li Z, Liang F, et al. Wavelet based human target detection in complex ruins using a low center frequency UWB radar[C]//2016 Progress in Electromagnetic Research Symposium (PIERS). IEEE, 2016: 1744-1747.

[59] QIU L, JIN T, LU B Y, et al. An isophase-based life signal extraction in through-the-wall radar[J]. IEEE Geoscience and Remote Sensing Letters,2016,14(2):193-197.

[60] 张崇超,张长春,张群英.EEMD 在超宽带雷达生命信号提取中的应用[J].电子测量技术,2012,35(4):76-80.

[61] XU Y Y, DAI S, WU S Y, et al. Vital sign detection method based on multiple higher order cumulant for ultrawideband radar[J]. IEEE Transactions on Geoscience and Remote Sensing,2011,50(4):1254-1265.

[62] LV H, JIAO T, ZHANG Y, et al. A novel method for breath detection via stepped-frequency continuous wave ultra-wideband(SFCW UWB)radars based on operational bandwidth segmentation[J].Sensors,2018,18(11):3873.

[63] 陈瑞鼎,鹿琪,单子涵,等.基于卡尔曼滤波的超宽带穿墙雷达移动目标探测[J].地球物理学进展,2017,32(4):1758-1763.

[64] 胡巍.基于多普勒雷达的非接触式生命体征检测技术研究[D].中国科学技术大学,2014.

[65] 周小龙.无载波超宽带雷达人体动作识别理论方法研究[D].桂林电子科技大学, 2019.

[66] YEH J R,SHIEH J S,HUANG N E. Complementary ensemble empirical mode decomposition: A novel noise enhanced data analysis method[J].Advances in adaptive data analysis,2010,2(02):135-156.

[67] Shyu Kuokai, Chiu Luanjiau, Lee Polei, et al. Detection of breathing and heart rates in UWB radar sensor date using FVPIEF-based two-layer EEMD[J].IEEE Sensors Journal, 2018,19,774-784.

[68] 崔学荣,杨雷,李娟,等.复杂场景下基于 UWB 雷达的呼吸特征检测算法[J].电子测量技术,2021,44(4):70-74.

[69] 刘超,高勇.K-L 变换应用于 UWB 雷达回波信号的增强[J].无线电工程,2017,47(7): 34-37.

[70] 郑学召,丁文,黄渊,等.不同领域下超宽带雷达探测呼吸心跳信号研究综述[J].雷达学报（中英文）,2025,14(1):204-228.

[71] 肖明国,张彪,康玉国,等.面向钻孔救援的 UWB 雷达探测技术研究进展[J].煤炭技术,2024,43(06):174-177.

[72] Taylor, James D., ed. Ultra-wideband radar technology. CRC press, 2018.

[73] 王中林,邵佳佳. 面向工程电磁学的动生麦克斯韦方程组及其求解方法[J].中国科学：技术科学,2022,52(9):1416-1433.

[74] 李蕙君,钟若飞. 探地雷达波振幅与土壤含水量关系的数值模拟[J].应用科学学报,2015,33(1):41-49.

[75] Schires E, Georgiou P, Lande T S. Vital sign monitoring through the back using an UWB impulse radar with body coupled antennas[J]. IEEE transactions on biomedical circuits and systems 2018,12(2): 292-302.

[76] 张怡.基于 UWB 雷达的多功能应用算法研究[D].北京邮电大学,2020.

[77] 李永甫,吴斌,包宇,等.四刚体人体行走模型及其人行荷载分析[J].振动工程学报,2022,35(04):903-911.

[78] 金添,何元,李新羽,等.超宽带雷达人体行为感知研究进展[J].电子与信息学报,2022,44(04):1147-1155.

[79] 刘翼群.生命探测雷达信号处理算法研究[D].西安电子科技大学,2019.

[80] 侯敏.基于深度学习的人体微动特性分析与步态识别[D].西安电子科技大学,20121.

[81] YANG Degui, ZHU Zhengliang, ZHANG Junchao, et al. The overview of human localization and vital sign signal measurement using handheld IR-UWB through-wall radar[J]. Sensors, 2021, 21(2): 402.

[82] 张铎.超宽带雷达波在煤中传播规律与定位基础研究[D].西安科技大学,2018.

[83] 徐向阳.基于 UWB 生物雷达的心跳呼吸提取技术实现[D].西安电子科技大学,2018.

[84] 朱少民,夏虹,尹文哲,等.基于变分模态分解和希尔伯特变换的转子非平稳信号故障特征识别[J].哈尔滨工程大学学报,2024,45(05):825-832.

[85] 杨路,雷雨霄,余翔.基于 FMCW 雷达的人体生命体征信号预测算法[J].雷达科学与技术,2024,22(01):43-56.

[86] 马铖.面向低信噪杂比环境的随机信号雷达生命探测方法研究[D].太原理工大学,2022.

[87] 李朝锋,王振,刘欣宇,等.基于 VMD-ICA 的水下激光雷达抗散射信号处理方法[J].物理学报,2024,73(09):91-96.

[88] 唐刚,徐智.基于 VMD 与 ICA 的地震信号去噪方法[C].中国地球物理学会,中国地震学会,全国岩石学与地球动力学研讨会组委会,中国地质学会构造地质学与地球动力学专业委员会,中国地质学会区域地质与成矿专业委员会.北京化工大学；2019:1.

[89] WANG Guangjun, WANG Zhaohui, JIANG Hui, et al. Noise Reduction Method of Transient Electromagnetic Signal Based on VMD-ICA[C]. 2021 40th Chinese Control Conference (CCC), Shanghai, China, 2021, pp. 3132-3137.

[90] 胡洁,张义平,郑禄林,等.基于围岩电性响应特征的钻孔雷达探测技术研究与应用[J].矿业研究与开发,2025,45(02):246-253.