基于FMCW 毫米波雷达的跌倒检测

跌倒行为是对老人身体健康的潜在威胁，据统计摔倒是导致无意识损伤的主
要原因。以更快侦测出老人摔倒行为并采取补救措施为目的，本文使用调频连续
波毫米波雷达设计实现了摔倒行为的高效检测。本文使用电磁仿真软件对雷达的
微带天线进行了仿真，获得了微带天线的增益和方向图，并结合雷达的工作原理
对仿真结果进行了分析。由于点云质量受天线的增益影响，本文对龙伯透镜进行
仿真，对比了天线的方向图，验证了使用透镜天线可以提高雷达天线峰值增益。本
文对雷达数据进行了处理并使用密度聚类算法和卡尔曼滤波算法实现了房间内的
人员识别和轨迹追踪。在实现摔倒行为检测中，本文结合了阈值检测和神经网络，
其中阈值检测通过目标的重心位置判断人是否摔倒，由变分自编码器和两种时间
序列模型构成的神经网络通过训练不包括跌倒的日常生活数据得到模型，所获模
型通过对比跌倒前后的损失函数实现摔倒行为甄别，最后成功地实现了高效的摔
倒行为检测。

Falling is the main underlying threat to old people, It is reported that falls are prominent cause of unintentional injuries. In order to quickly detect the fall of the elderly and promptly take remedial measures, this paper uses the FMCW millimeter wave radar to achieve efficient fall detections. Firstly the electromagnetic simulation software HFSS is used to simulate the microstrip antenna of the radar, the gain and direction of the microstrip antenna are obtained, the simulation results are analyzed in combination with the working principle of the radar. Since the point cloud quality is affected by the gain of the antenna, this paper simulates the Lunberg lens and compares the direction diagram of the antenna to verify that the peak gain of the radar antenna can be improved by using the lens antenna. Finally, this paper processes the radar data and uses the DBSCAN clustering algorithm and kalman filtering algorithm to realize the person identification and the trajectory tracking in a room. In the fall detection, this paper also successfully combines threshold detection with neural network, in which the threshold detection decides whether a person falls by estimating the position of the center of gravity, and the neural network is compose of variational autoencoder and two kinds of sequence-to-sequence network models which reduces the loss by training the daily life data that does not include falls. Thus the obtained model realizes the fall behavior distinction by comparing the loss before and after the fall. To conclude, the combination of the threshold detection and the neural network training have successfully achieved the high efficiency of fall detections.

[1] 刘阳, 戴泽阳, 蒋晗, 等. 江苏省不同年龄段老年人养老意愿及影响因素分析[J]. 重庆医学, 2021.
[2] WORLD HEALTH ORGANIZATION. WHO global report on falls prevention in older age[M]. World Health Organization, 2008.
[3] ALEXANDER B H, RIVARA F P, WOLF M E. The cost and frequency of hospitalization for fall-related injuries in older adults[J]. American Journal of Public Health, 1992, 82(7): 1020-1023.
[4] 尹香君, 施小明, 司向, 等. 中国疾病预防控制系统慢性非传染性疾病预防控制能力评估[J]. 中华流行病学杂志, 2010(10): 1125-1129.
[5] CHENG P, WANG L, NING P, et al. Unintentional falls mortality in China, 2006-2016[J].Journal of Global Health, 2019, 9(1).
[6] SPOSARO F, TYSON G. IFALL: An android application for fall monitoring and response[C]//2009 Annual International Conference of the IEEE Engineering in Medicine and Biology Society. 2009: 6119-6122.
[7] WANG X, ELLUL J, AZZOPARDI G. Elderly fall detection systems: A literature survey[J].Frontiers in Robotics and AI, 2020, 7: 71.
[8] BHATTACHARYA A, VAUGHAN R. Deep learning radar design for breathing and fall detection[J]. IEEE Sensors Journal, 2020, 20(9): 5072-5085.
[9] TEXAS INSTRUMENTS. The fundamentals of millimeter wave radar sensors[EB/OL]. 2020.https://www.ti.com/lit/wp/spyy005a/spyy005a.pdf.
[10] IGUAL R, MEDRANO C, PLAZA I. Challenges issues and trends in fall detection systems[J].BioMedical Engineering OnLine, 2013, 12: 66.
[11] MERCURI M, SCHREURS D, LEROUX P. SFCW microwave radar for in-door fall detection[C]//2012 IEEE Topical Conference on Biomedical Wireless Technologies, Networks, and Sensing Systems (BioWireleSS). 2012: 53-56.
[12] MOENESS G A. Micro-doppler classification of activities of daily living incorporating human ethogram[C]//Proc.SPIE: volume 11408. 2020.
[13] LIU L, POPESCU M, SKUBIC M, et al. Automatic fall detection based on Doppler radar motion signature[C]//2011 5th International Conference on Pervasive Computing Technologies for Healthcare (PervasiveHealth) and Workshops. IEEE, 2011: 222-225.
[14] WU Q, ZHANG Y D, TAO W, et al. Radar-based fall detection based on doppler time–frequency signatures for assisted living[J]. IET Radar, Sonar & Navigation, 2015, 9(2): 164-172.
[15] AMIN M G, ZHANG Y D, AHMAD F, et al. Radar signal processing for elderly fall detection:the future for in-home monitoring[J]. IEEE Signal Processing Magazine, 2016, 33(2): 71-80.
[16] JOKANOVIC B, AMIN M, AHMAD F, et al. Radar fall detection using principal component analysis[C]//Radar Sensor Technology XX: volume 9829. SPIE, 2016: 358-363.
[17] ZHAO P, LU C X, WANG J, et al. MID: tracking and identifying people with millimeter wave radar[C]//2019 15th International Conference on Distributed Computing in Sensor Systems(DCOSS). 2019: 33-40.
[18] ZHAO M, TIAN Y, ZHAO H, et al. RF-based 3D skeletons[C]//SIGCOMM ’18: Proceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication. New York, NY, USA: Association for Computing Machinery, 2018: 267–281.
[19] WANG B, GUO L, ZHANG H, et al. A millimetre-wave radar-based fall detection method using line kernel convolutional neural network[J]. IEEE Sensors Journal, 2020, 20(22): 13364-13370.
[20] TEXAS INSTRUMENTS. MIMO radar[EB/OL]. 2017. https://www.ti.com/lit/an/swra554a/swra554a.pdf.
[21] JIN F, SENGUPTA A, CAO S. MMFall: Fall Detection Using 4-D mmWave Radar and a Hybrid Variational RNN AutoEncoder[J]. IEEE Transactions on Automation Science and Engineering,2022, 19(2): 1245-1257.
[22] MEDSKER L, JAIN L C. Recurrent neural networks: design and applications[M]. CRC press,1999.
[23] CHO K, VAN MERRIENBOER B, BAHDANAU D, et al. On the properties of neural machine translation: encoder-decoder approaches[M/OL]. ArXiv, 2014. https://arxiv.org/abs/1409.1259.
[24] 苏剑林. 也来谈谈RNN 的梯度消失/爆炸问题[EB/OL]. 2020. https://kexue.fm/archives/7888.
[25] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation,1997, 9(8): 1735-1780.
[26] KRAMER M A. Nonlinear principal component analysis using autoassociative neural networks[J]. AIChE Journal, 1991, 37(2): 233-243.
[27] HOTELLING H. Analysis of a complex of statistical variables into principal components.[J].Journal of Educational Psychology, 1933, 24(6): 417.
[28] YU G, SAPIRO G, MALLAT S. Solving inverse problems with piecewise linear estimators:From Gaussian mixture models to structured sparsity[J]. IEEE Transactions on Image Processing,2011, 21(5): 2481-2499.
[29] KINGMA D P, WELLING M. Auto-encoding variational bayes[M/OL]. arXiv, 2013. https://arxiv.org/abs/1312.6114.
[30] 苏剑林. 变分自编码器（一）：原来是这么一回事[EB/OL]. 2018. https://kexue.fm/archives/5253.
[31] 宋旭亮. 矩形微带天线设计与阻抗匹配网络[D]. 大连: 大连海事大学, 2008.
[32] 李明洋，刘敏，杨放. HFSS 天线设计[M]. 电子工业出版社, 2011.
[33] 王坤鹏. 车载毫米波防撞雷达天线的研究与设计[D]. 西安: 西安电子科技大学, 2011.
[34] QIU J, GUO X, QIU S, et al. A design of 94GHz millimeter wave individual soldier array antenna[C]//2019 IEEE Asia-Pacific Microwave Conference (APMC). 2019: 252-254.
[35] ROGERS CORPORATION. RO4835™ laminate data sheet[EB/OL]. 2020. https://rogerscorp.com/-/media/project/rogerscorp/documents/advanced-connectivity-solutions/english/data-sheets/ro4835-laminate-data-sheet.pdf.
[36] TEXAS INSTRUMENTS. 60GHz mmWave sensor evms[EB/OL]. 2020. https://www.ti.com/lit/ug/swru546d/swru546d.pdf.
[37] 王剑辉. 毫米波车载雷达天线研究[D]. 南京: 南京邮电大学, 2021.
[38] 王存. 高增益宽角扫描透镜天线及其阵列[D]. 合肥: 中国科学技术大学, 2020.
[39] DU G, LIANG M, SABORY-GARCIA R A, et al. 3D printing implementation of an X-band eaton lens for beam deflection[J]. IEEE Antennas and Wireless Propagation Letters, 2016, 15:1487-1490.
[40] 童文浩. 基于雷达的室内运动目标识别与跟踪系统的设计与实现[D]. 武汉: 华中科技大学, 2019.
[41] TEXAS INSTRUMENTS. XWR1843 evaluation module (xWR1843BOOST) single-chip mmWave sensing solution[EB/OL]. 2019. https://www.ti.com/lit/ug/spruim4b/spruim4b.pdf.
[42] MACQUEEN J, et al. Some methods for classification and analysis of multivariate observations[C]//Proceedings of the fifth Berkeley symposium on mathematical statistics and probability:volume 1. Oakland, CA, USA, 1967: 281-297.
[43] PEDREGOSA F, VAROQUAUX G, GRAMFORT A, et al. Scikit-learn: machine learning in python[J]. Journal of Machine Learning Research, 2011, 12: 2825-2830.
[44] ESTER M, KRIEGEL H P, SANDER J, et al. A density-based algorithm for discovering clusters in large spatial databases with noise.[C]//kdd: volume 96. 1996: 226-231.
[45] HUANG X, CHEENA H, THOMAS A, et al. Indoor detection and tracking of people using mmWave sensor[J]. Journal of Sensors, 2021, 2021: 6657709.
[46] ZARCHAN P. Progress in astronautics and aeronautics: fundamentals of kalman filtering: a practical approach: volume 208[M]. Aiaa, 2005.
[47] KALMAN R E. A new approach to linear filtering and prediction problems[J]. Transactions of the ASME–Journal of Basic Engineering, 1960, 82(Series D): 35-45.
[48] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[J]. Advances in Neural Information Processing Systems, 2017, 30: 6000-6010.
[49] ZHAO H, JIANG L, JIA J, et al. Point transformer[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. 2021: 16259-16268.