本論文旨在開發可自動化偵測睡眠活動之演算法，此演算法使用可攜式加速度訊號感測模組蒐集加速度訊號，並利用加速度計內部的有限狀態機將原始訊號轉換成特徵，透過藍芽的溝通將儲存之資料傳輸至雲端的睡眠活動偵測演算法進行運算，分析資料中的睡眠活動。睡眠活動偵測演算法分為三部分，其中的動靜態活動偵測演算法會將資料先分為靜態及動態活動，所有的靜態活動可能包含睡眠活動，透過睡眠活動偵測演算法區分睡眠活動與ㄧ般靜態活動，再經由例外排除演算法進行檢查，最終輸出睡眠起始時間、結束時間以及睡眠總時間。為了提高準確率，本論文開發之睡眠活動偵測演算法其使用的參數分別採用粒子群演算法及基因演算法進行參數最佳化。以過零值法為特徵，經過參數最佳化後之睡眠起始時間偵測準確率為81.67%，而睡眠結束時間偵測準確率則為88.33%;以過臨界值法產生之特徵，睡眠起始時間偵測準確率為81.67%，睡眠結束時間偵測準確率則為90.00%。本論文針對實驗結果進行討論，包含誤差的分布、產生的原因以及錯誤偵測睡眠活動的原因，並提出改善方式。研究結果證明了本演算法應用於自動化偵測睡眠活動的可行性，並在最後提出可能的改善方式，以進一步的增進睡眠活動偵測演算法的準確率並將此演算法實際應用於生活中。

This thesis proposes an automatic sleep activity detection algorithm for a wearable device which can detect sleep activity in the daily life. The wearable device contains a triaxial accelerometer embedded with a finite state machine, a bluetooth module with a microcontroller, to collect the signals for sleep activity detection. The proposed algorithm can detect sleep activities, including the on-bed time, the off-bed time, and the total sleep time. To improve the performance of the algorithm, the particle swarm optimization algorithm (PSO) and gene optimization algorithm (GA) are used to optimize the parameters of the algorithm. The experimental results show that the best accuracy of the on-bed time using the zero-crossing mode is 81.67%, and the best accuracy of the off-bed time is 88.33%. The best accuracy of the on-bed time and the off-bed time using the time-above-threshold mode is 81.67% and 90.00%, respectively. The results validate the efficiency of the proposed algorithm. Finally, the results are discussed, including the error distribution and the interpretation for sleep activity detection error. Suggestions are provided to further enhance the accuracy of the sleep activity detection algorithm for daily life applications.

[1]S. T. Aaronson, S. Rashed, M. P. Biber, J. A. Hobson, “Brain state and body position: A time-lapse video study of sleep,” Archives of General Psychiatry, vol. 39, no. 3, pp. 330-335, 1982.
[2]Y. Bao, Z. Hu, and T. Xiong, “A PSO and pattern search based memetic algorithm for SVMs parameters optimization,” Neurocomputing, vol. 117, no. 6, 2013.
[3]R. J. Cole, D. F. Kripke, D. J. Mullaney, and J. C. Gillin, “Technical note automatic sleep/wake identification from wrist activity,” Sleep, vol. 15, no. 55, pp. 461-469, 1992.
[4]D. F. Cook, C. T. Ragsdale, and R. L. Major, “Combining a neural network with a genetic algorithm for process parameter optimization,” Engineering Applications of Artificial Intelligence, vol. 13, no. 4, 2000.
[5]Y. Cheng, C. L. Du, J. J. Hwang, I. S. Chen, M. F. Chen, and T. C. Su, “Working hours, sleep duration and the risk of acute coronary heart disease: A case–control study of middle-aged men in Taiwan,” International Journal of Cardiology, vol. 171, no. 3, pp. 419-422, 2014.
[6]K. Deep, and M. Thakur, “A new mutation operator for real coded genetic algorithms,” Applied Mathematics and Computation, vol. 193, no. 1, 2007.
[7]R. C. Eberhart and J. Kennedy, “A new optimizer using particle swarm theory,” In Proceedings of the sixth International Symposium on Micro Machine and Human Science, pp. 39-43, 1995.
[8]V. G. Gudise, and K. Ganesh, “Comparison of particle swarm optimization and backpropagation as training algorithms for neural networks,” Swarm Intelligence Symposium, 2003.
[9]J. H. Holland, “Adaptation in natural and artificial systems,” The University of Michingan Press, 1975.
[10]J. Horne, “Short sleep is a questionable risk factor for obesity and related disorders: Statistical versus clinical significance,” Biological Psychology, vol. 77, no. 3, pp. 266-276, 2008.
[11]M. Irwin, J. Mcclintick, C. Costlow, M. Fortner, J. White, and J. C. Gillin, “Partial night sleep deprivation reduces natural killer and cellular immune responses in humans,” The FASEB journal, vol. 10, no. 5, pp. 643-653, 1996.
[12]G. Jean-Louis, D. F. Kripke, R. J. Cole, J. D. Assmus, and R. D. Langer, “Sleep detection with an accelerometer actigraph comparisons,” Physiology & Behavior, vol. 72, no. 1, pp. 21-28, 2001.
[13]G. Jean-Louis, D. F. Kripke, W. J. Mason, J. A. Elliott, and S. D. Youngstedt, “Sleep estimation from wrist movement quantified by different actigraphic modalities,” Journal of Neuroscience methods, vol. 105, no. 2, pp. 185-191, 2001.
[14]S. Javaheri, A. Storfer-Isse, C. L. Rosen, and S. Redline, “Sleep quality and elevated blood pressure in adolescents,” Circulation, vol. 118, no. 10, pp. 1034-1040, 2008.
[15]M. F. Jjorth, J. P. Chaput, C. T. Damsgaard, S. M. Dalskov, K. F. Michaelsen, I. Tetens, and A. Sjodin, “Measure of sleep and physical activity by a single accelerometer: Can a waist‐worn actigraph adequately measure sleep in children,” Sleep and Biological Rhythms, vol. 10, no. 4, pp. 328-335, 2012.
[16]J. R. Kinder, K. A. Lee, H. Thompson, K. Hicks, K. Topp, and K. A. Madsen, “Validation of a hip-worn accelerometer in measuring sleep time in children,” Journal of Pediatric Nursing, no.27, vol. 2, pp. 127-133, 2012.
[17]C. Lin, and M. H. Hsieh, “Classification of mental task from EEG data using neural networks based on particle swarm optimization,” Neurocomputing, vol. 72, no. 4, pp. 1121-1130, 2009.
[18]S. R. Patel and F. B. Hu, “Short sleep duration and weight gain: A systematic review,” Obesity, vol. 16, no. 3, pp. 643–653, 2008.
[19]V. Pillai, L. A. Steenburg, J. A. Ciesla, T. Roth, and , C. L. Drake, “A seven day actigraphy-based study of rumination and sleep disturbance among young adults with depressive symptoms,” Jounal of Psychosomatic Research, vol. 77, no. 1, pp. 70-75, 2014.
[20]A. Sadeh, K. M. Sharkey, and M. A. Carskadon, “Activity-based sleep—wake identification: An empirical test of methodological Issues,” Sleep, vol. 17, no. 3, pp. 201-207, 1994.
[21]A. Sadeh, P. J. Hauri, D. F. Kripke, P. Lavie, “The role of actigraphy in the evaluation of sleep disorders,” Sleep, vol. 18, no. 4, pp. 228-302, 1995.
[22]A. M. Swartz, S. J. Strath, D. R. Bassett, W. L. O’brien, G. A. King, and B. E. Ainsworth, “Objective sleep measures and subjective sleep satisfaction: How do older adults with insomnia define a good night’s sleep,” Psychology and Aging, vol. 13, no. 1, pp. 159-163, 1998.
[23]A. Sadeh, “The role and validity of actigraphy in sleep medicine,” Sleep Medicine Reviews, vol. 15, no. 4, pp. 259-267, 2011.
[24]A. Vallieres, and C. M. Morin, “Actigraphy in the assessment of insomnia,” Sleep-new York Then Westchester, vol. 26 no. 7, pp. 902-906, 2003.
[25]J. F. Van Den Berg, F. J. Van Rooij, H. Vos, J. H. Tulen, A. Hofman, H. M. Miedema, A. K. Neven, and H. Tiemeier, “Disagreement between subjective and actigraphic measures of sleep duration in a population-based study of elderly persons,” Journal of Sleep Research, vol. 17, no. 3, pp. 295-302, 2008.
[26]A. M. Williamson, and A. M. Feyer, “Moderate sleep deprivation produces impairments in cognitive and motor performance equivalent to legally prescribed levels of alcohol intoxication,” Occupational and Environmental Medicine, vol. 57, no. 10, pp. 649-655, 2000.
[27]W. C. Wu, and M. S. Tsai. “Application of enhanced integer coded particle swarm optimization for distribution system feeder reconfiguration,” IEEE Transactions on Power System, vol. 26, no. 3, pp. 1591-1599, 2011.
[28]Online available: http://www.theactigraph.com
[29]Online available: https://www.fitbit.com/cn
[30]Online available: http://www.mi.com/tw/miband/
[31]蘇木春、張孝德，機器學習：類神經網路、模糊系統以及基因演算法則，全華，台北，民93。
[32]王進德，類神經網路與模糊控制理論：入門與應用，全華，台北，民95。
[33]CC2541 data sheet, Texas Instruments.
[34]LIS3DSH data sheet, ST Microelectronics.
[35]LIS3DSH based Pedometer3, ST Microelectronics.