人體靜止狀態下的呼吸率測量，不論是接觸式感測技術或非接觸式感測技術都有相當高的準確度，但是對於移動中的人體呼吸率感測卻是一大挑戰。因此本論文結合非接觸和接觸式感測技術，將非接觸式的頻率調變連續波雷達放至人體手腕處進行實驗，不僅能降低傳統將呼吸帶戴在胸口上的不適感，更能偵測到人體運動中的呼吸率。
在過去認為呼吸率只能透過胸腔的起伏獲得，本論文透過實驗證明當人體呼吸時，人體筋膜會受到拉扯，因此將雷達放在手部任意處都能夠透過筋膜的變化獲得呼吸訊號。
對於移動人體呼吸率的感測，本論文提出一套訊號處理流程，首先，對雷達輸出的訊號進行解調後，透過離散小波轉換和獨立成分分析兩種方式來分解重建生理訊號，得到呼吸的頻率。當人體移動時，由於有許多雜訊摻雜在訊號中，因此離散小波轉換所分解的訊號中含有多個頻率點，導致我們在選擇分量時無從選擇。透過獨立成分分析能夠解決選擇分量的問題，也能直接得出呼吸率。本論文整理了這兩種方法的效能評估，獨立成分分析演算法對於呼吸率偵測的均方根誤差為每分鐘2.784下，而離散小波轉換和獨立成分分析的結合演算法為每分鐘3.239下。

Whether it is contact sensing technology or non-contact sensing technology, the respiration rate measurement of the human body in a static state has a very high accuracy, but sensing the respiration rate of the moving human body is a big challenge. Therefore, this paper combines non-contact and contact sensing technology and puts the non-contact frequency modulated continuous wave radar on the human wrist for experiments. It can not only reduce the discomfort of wearing the traditional breathing belt on the chest, but also detect measurement of human respiration rate during movement.
In the past, it was thought that the respiratory rate could only be obtained through the ups and downs of the chest. This paper proves through experiments that when the human body breathes, the human fascia will be stretched. Therefore, placing the radar anywhere on the hand can obtain the respiration through the changes in the fascia.
For sensing the breathing rate of a moving human body, this paper proposes a set of signal processing processes. First, after demodulating the radar output signal, the physiological signal is decomposed and reconstructed through discrete wavelet transformation and independent component analysis to obtain the respiratory rate. frequency. When the human body moves, because there is a lot of noise mixed in the signal, the signal decomposed by discrete wavelet transformation contains multiple frequency points, leaving us with no choice when selecting components. The problem of component selection can be solved through independent component analysis, and the respiratory rate can also be obtained directly. This paper summarizes the performance evaluation of these two methods. The root mean square error of the independent component analysis algorithm for respiratory rate detection is 2.784 beats per minute, while the combined algorithm of discrete wavelet transform and independent component analysis is 3.239 beats per minute.

論文審定書.................................................................................................................................................. i
誌謝...............................................................................................................................................................ii
摘要.............................................................................................................................................................. iii
Abstract........................................................................................................................................................ iv
目錄.............................................................................................................................................................. vi
圖次.............................................................................................................................................................. viii
表次.............................................................................................................................................................. x
第一章 序論................................................................................................................................................. 1
1.1 研究背景與動機.................................................................................................................................... 1
1.2 頻率調變連續波雷達............................................................................................................................ 3
1.3 演算法回顧............................................................................................................................................ 5
1.3.1 小波轉換............................................................................................................................................. 5
1.3.2 獨立成分分析..................................................................................................................................... 6
1.4 章節規劃................................................................................................................................................ 7
第二章 系統架構與演算法理論................................................................................................................. 9
2.1 頻率調變連續波雷達解調.................................................................................................................... 9
2.1.1 理想環境下之拍頻相位分析......................................................................................................... 9
2.1.2 空間中存在非穩態雜波下之拍頻相位分析................................................................................. 10
2.2 小波轉換.................................................................................................................................................11
2.2.1 離散小波轉換..................................................................................................................................12
2.2.2 最大重疊離散小波轉換..................................................................................................................14
2.3 獨立成分分析 .......................................................................................................................................16
第三章 筋膜實驗與演算法結果................................................................................................................. 20
3.1 實驗設置................................................................................................................................................ 20
3.2 演算法介紹與筋膜實驗結果................................................................................................................ 24
3.2.1 演算法介紹..................................................................................................................................... 24
3.2.2 筋膜實驗結果................................................................................................................................. 25
3.3 小波轉換VS. 其他訊號分解法............................................................................................................. 29
第四章 生理感測實驗................................................................................................................................. 36
4.1 雷達實驗設置........................................................................................................................................ 36
4.2 生理感測實驗結果............................................................................................................................... 37
4.2.1 獨立成分分析結果......................................................................................................................... 37
4.2.2 獨立成分分析+小波轉換結果....................................................................................................... 44
4.3 效能比較................................................................................................................................................ 48
第五章 結論................................................................................................................................................. 52
參考文獻...................................................................................................................................................... 53

[1] N. Daimiwal, M. Sundhararajan, and R. Shriram, “Respiratory rate, heart rate and continuous measurement of BP using PPG,” in Proc. Int. Conf. Commun. Signal Process., 2014, pp. 999–1002. [Online]. Available: http://doi.org/10.1109/ICCSP.2014.6949996
[2] M. Nosrati and N. Tavassolian, “High-accuracy heart rate variability monitoring using Doppler radar based on Gaussian pulse train modeling and FTPR algorithm,” IEEE Trans. Microw. Theory Techn., vol. 66, no. 1, pp. 556-567, Jan. 2018.
[3] W. Massagram, V. M. Lubecke, A. Host-Madsten, O. BoricLubecke, “Assessment of Heart Rate Variability and Respiratory Sinus Arrythmia via Doppler Radar,” IEEE Trans. Microw. Theory Techn., vol. 57, no. 10, pp. 2542-2549, Oct. 2009.
[4] Chen Ye, Kentaroh Toyoda, Tomoaki Ohtsuki, “A Stochastic Gradient Approach for Robust Heartbeat Detection With Doppler Radar Using Time-Window-Variation Technique,” IEEE Trans. Biomed. Eng., vol.66, no.6, pp.1730-1741, Jun. 2019.
[5] V. L. Petrovic, M. M. Jankovic, A. V. Lupsic, V. R. Mihajlovic, and J. S. Popovic-Bozovic, “High-accuracy real-time monitoring of heart rate variability using 24 GHz continuous-wave Doppler radar,” IEEE Access, vol. 7, pp. 74721-74733, Jan. 2019.
[6] HRV[Online].Available:https://cdntwrunning.biji.co/808_77d442ea9e54a90a9c10d9760eceec85a5e00132373a7031000fd987a3c9f87b.jpg
[7] C. Li, V. M. Lubecke, O. Boric-Lubecke, and J. Lin, “A review on recent advances in doppler radar sensors for noncontact healthcare monitoring,” IEEE Trans. Microw. Theory Techn., vol. 61, no. 5, pp. 2046-2060, May 2013.
[8] A. Al-Naji, K. Gibson, S.-H. Lee and J. Chahl, “Monitoring of cardiorespiratory signal: Principles of remote measurements and review of methods,” IEEE Access, vol. 5, pp. 15776-15790, 2017.
[9] F.-K. Wang et al., “Review of self-injection-locked radar systems for noncontact detection of vital signs,” IEEE J. Electromagn., RF Microw. Med. Biol., vol. 4, no. 4, pp. 294-307, Dec. 2020.
[10] Muñoz-Ferreras, J.M.; Gómez-García, R.; Li, C. Chapter 5—Human-aware localization using linear-frequency-modulated continuous-wave radars. In Principles and Applications of RF/Microwave in Healthcare and Biosensing; Li, C., Tofighi, M.R., Schreurs, D., Horng, T.S.J., Eds.; Academic Press: Cambridge, MA, USA, 2017; pp. 191–242.
[11] W. Hu, Z. Zhao, Y. Wang, H. Zhang and F. Lin, “Noncontact Accurate Measurement of Cardiopulmonary Activity Using a Compact Quadrature Doppler Radar Sensor,” IEEE Trans. Biomed. Eng., vol. 61, no. 3, pp. 725-735, Mar. 2014.
[12] ICA和fastICA演算法[Online].Available:https://learning-archive.org/wp-content/uploads/2022/07/%E7%8B%AC%E7%AB%8B%E6%88%90%E5%88%86%E5%88%86%E6%9E%90%E5%92%8CFastICA%E7%AE%97%E6%B3%95.pdf
[13] ICA理論[Online].Available:https://tomohiroliu22.medium.com/%E6%A9%9F%E5%99%A8%E5%AD%B8%E7%BF%92-%E5%AD%B8%E7%BF%92%E7%AD%86%E8%A8%98%E7%B3%BB%E5%88%97-64-%E7%8D%A8%E7%AB%8B%E6%88%90%E5%88%86%E5%88%86%E6%9E%90-independent-component-analysis-53bc1a530146
[14] P. Nallabolu, L. Zhang, H. Hong, and C. Li, “Human presence sensing and gesture recognition for smart home applications with moving and stationary clutter suppression using a 60-GHz digital beamforming FMCW radar,” IEEE Access, vol. 9, pp. 72857–72866, 2021.
[15] D. B. Percival and A. T. Walden, “Wavelet methods for time series analysis (cambridge series in statistical and probabilistic mathematics),” 2000.
[16] 筋膜系統[Online].Available:https://www.vghtc.gov.tw/UploadFiles/Images/EDM/2b982258-beea-4d49-89b3-47e77bc45efe_2.pdf
[17] 橫膈膜如何影響筋膜系統[Online].Available:https://www.vghtc.gov.tw/UploadFiles/WebFiles/WebPagesFiles/Files/9063d5ce-9c27-4bcf-93e7-1a0844a6436b/%E6%A9%AB%E8%86%88%E8%86%9C%E7%9A%84%E9%9B%99%E9%87%8D%E5%8A%9F%E8%83%BD%EF%BC%8D%E5%91%BC%E5%90%B8%E3%80%81%E5%A7%BF%E5%8B%A2%E6%A0%B8%E5%BF%83%E7%A9%A9%E5%AE%9A.pdf
[18] 人體手臂筋膜圖[Online].https://daybreakphysio.com/know_more_about_fascia_part_one/
[19] Infineon[Online].Available:https://www.infineon.com/cms/en/product/sensor/radar-sensors/radar-sensors-for-iot/60ghz-radar/bgt60tr13c/#!documents
[20] 星宇儀器有限公司「Go Direct 呼吸監控胸帶」[Online].Available:https://www.hy6688.com.tw/product-detail-2976951.html