負重聳肩跳、負重懸垂式上搏與反向跳皆為目前運動員常見訓練項目，主要用於訓練或測試瞬時爆發力，執行負重懸垂式上搏之過程中需要保證動作順暢協調的同時，以正確的姿勢完成動作皆為主要目標。近年來基於人工神經網路的發展，原本需要大量程式分析與數學模型的運算才能取得之生物力學數值，可以透過訓練神經網路來減少分析成本與時間消耗。
　　因此本研究以人工神經網路為開發基礎，主要圍繞於膝關節角度、踝關節角度與髖關節角度作為主要探討部分。並以對於負重聳肩跳、負重懸垂式上搏與反向跳有興趣的研究參與者進行測試與資料收集，在各種負重狀態進行聳肩跳與負重懸垂式上搏運動與無負重反向跳，藉由三維動作捕捉系統收集之資料進行運算得出關節角實際數值，並結合慣性測量單元數據，給予人工神經網路進行學習與預測。
　　根據本研究所產生之神經網路預測結果可以發現，在重量訓練動作預測中其相對誤差處於可以接受的範圍，在留一驗證中相對均方根誤差範圍最高平均為15.2819%，最高平均誤差為7.8692度，與真實資料相關係數皆大於0.85，絕對誤差於3.41至9.3度之內，使其可以作為初步分析與回饋系統使用，並且為使其用於實際重量訓練中，驗證過程亦使用獨立測試集進行驗證與評估，其獨立驗證結果中相對均方根誤差範圍最高來至19.587%，最高的平均誤差度數則來到了11.8024度，雖誤差相較留一驗證高，但與真實資料之相關係數亦大於0.85，也展現其針對未知資料實際應用能力，有助於在未來擴增資料庫數量與資料類型後獲得更準確之預測。

Jump shrug, hang power clean and countermovement jump are currently mainstream training items for athletes, primarily used for training or testing instantaneous explosive power. During the execution of hang power clean, it is necessary to ensuring smooth and coordinated movement, while completing the action with correct posture is the primary goal. In recent years, with the development of artificial neural networks, biomechanical values that previously required extensive programming analysis and mathematical model calculations can now be obtained through trained neural networks, reducing analysis costs and time consumption. This study will develop an artificial neural network joint angle conversion system suitable for jump shrug, hang power clean and countermovement jump to monitor movement precision and joint status while reducing sports injuries, equipment, analysis, and training costs.
　Therefore, this study was based on the mainstream artificial neural networks used in various literatures, mainly focusing on knee joint angles, ankle joint angles, and hip joint angles. It involved testing and data collection on subjects interested in jump shrug, hang power clean and countermovement jump, performing jump shrug and hang power clean movements under various weight-bearing conditions and unweighted countermovement jump. Joint angle actual values were calculated using data collected by a three-dimensional motion capture system and combined with inertial measurement unit data, provided to the artificial neural network for learning and prediction.
　The neural network prediction results from this study show relative errors within an acceptable range for predicting weight training movements, with an average root mean square error of 15.2819%. The highest average error is 7.8692 degrees, and correlation coefficients with real data are all greater than 0.85. Absolute errors range from 3.41 to 9.3 degrees, making it suitable for preliminary analysis and feedback systems in actual weight training. This system can also be expanded in the future to include more mechanical indicators, constructing a more comprehensive and multifunctional system to meet training and even academic research needs.

Adesida, Y., Papi, E., & McGregor, A. H. (2019). Exploring the Role of Wearable Technology in Sport Kinematics and Kinetics: A Systematic Review. Sensors, 19(7), 1597. Retrieved from https://www.mdpi.com/1424-8220/19/7/1597
Akoglu, H. (2018). User's guide to correlation coefficients. Turkish Journal of Emergency Medicine, 18(3), 91-93. doi:https://doi.org/10.1016/j.tjem.2018.08.001
Ancillao, A., Tedesco, S., Barton, J., & O'Flynn, B. (2018). Indirect Measurement of Ground Reaction Forces and Moments by Means of Wearable Inertial Sensors: A Systematic Review. Sensors (Basel), 18(8). doi:10.3390/s18082564
Bobbert, M. F., Gerritsen, K. G., Litjens, M. C., & Van Soest, A. J. (1996). Why is countermovement jump height greater than squat jump height? Medicine and science in sports and exercise, 28, 1402-1412.
Cloete, T., & Scheffer, C. (2008). Benchmarking of a full-body inertial motion capture system for clinical gait analysis. Paper presented at the 2008 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.
Comfort, P., & McMahon, J. J. (2015). Reliability of Maximal Back Squat and Power Clean Performances in Inexperienced Athletes. The Journal of Strength & Conditioning Research, 29(11), 3089-3096. doi:10.1519/jsc.0000000000000815
Conforti, I., Mileti, I., Del Prete, Z., & Palermo, E. (2020). Measuring Biomechanical Risk in Lifting Load Tasks Through Wearable System and Machine-Learning Approach. Sensors, 20(6), 1557. Retrieved from https://www.mdpi.com/1424-8220/20/6/1557
de Almeida, T. F., Morya, E., Rodrigues, A. C., & de Azevedo Dantas, A. F. O. (2021). Development of a Low-Cost Open-Source Measurement System for Joint Angle Estimation. Sensors (Basel), 21(19). doi:10.3390/s21196477
Duba, J., Kraemer, W. J., & Martin, G. (2007). A 6-step progression model for teaching the hang power clean. Strength & Conditioning Journal, 29(5), 26-35.
Dwork, C., Feldman, V., Hardt, M., Pitassi, T., Reingold, O., & Roth, A. (2015). Generalization in Adaptive Data Analysis and Holdout Reuse. https://proceedings.neurips.cc/paper_files/paper/2015/file/bad5f33780c42f2588878a9d07405083-Paper.pdf
Faigenbaum, A. D., McFarland, J. E., Herman, R., Naclerio, F., Ratamess, N. A., Kang, J., & Myer, G. D. (2012). Reliability of the one repetition-maximum power clean test in adolescent athletes. Journal of strength and conditioning research/National Strength & Conditioning Association, 26(2), 432.
Findlow, A., Goulermas, J. Y., Nester, C., Howard, D., & Kenney, L. P. J. (2008). Predicting lower limb joint kinematics using wearable motion sensors. Gait & Posture, 28(1), 120-126. doi:https://doi.org/10.1016/j.gaitpost.2007.11.001
Garhammer, J. (1984). Power Clean: Kinesiological evaluation. Strength & Conditioning Journal, 6(3), 40-40. Retrieved from https://journals.lww.com/nsca-scj/Fulltext/1984/06000/Power_Clean__Kinesiological_evaluation.7.aspx
Gurchiek, R. D., Cheney, N., & McGinnis, R. S. (2019). Estimating Biomechanical Time-Series with Wearable Sensors: A Systematic Review of Machine Learning Techniques. Sensors, 19(23), 5227. Retrieved from https://www.mdpi.com/1424-8220/19/23/5227
Hedrick, A., & Wada, H. (2008). Weightlifting movements: do the benefits outweigh the risks? Strength & Conditioning Journal, 30(6), 26-35.
Hori, N., Newton, R. U., Nosaka, K., & Stone, M. H. (2005). Weightlifting exercises enhance athletic performance that requires high-load speed strength. Strength & Conditioning Journal, 27(4), 50-55.
Hydock, D. (2001). The Weightlifting Pull in Power Development. Strength & Conditioning Journal, 23(1), 32. Retrieved from https://journals.lww.com/nsca-scj/fulltext/2001/02000/the_weightlifting_pull_in_power_development.6.aspx
Islam, M., Chen, G., & Jin, S. (2019). An overview of neural network. American Journal of Neural Networks and Applications, 5(1), 7-11.
Karatsidis, A., Bellusci, G., Schepers, H. M., De Zee, M., Andersen, M. S., & Veltink, P. H. (2017). Estimation of Ground Reaction Forces and Moments During Gait Using Only Inertial Motion Capture. Sensors, 17(1), 75. Retrieved from https://www.mdpi.com/1424-8220/17/1/75
Kingma, D. P., & Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
Kipp, K., Reddden, J., Sabick, M., & Harris, C. (2011). Weightlifting Performance Is Related to Kinematic and Kinetic Patterns of the Hip and Knee Joints. Journal of strength and conditioning research / National Strength & Conditioning Association, 26, 1838-1844. doi:10.1519/JSC.0b013e318239c1d2
Kobsar, D., Charlton, J. M., Tse, C. T., Esculier, J.-F., Graffos, A., Krowchuk, N. M., . . . Hunt, M. A. (2020). Validity and reliability of wearable inertial sensors in healthy adult walking: A systematic review and meta-analysis. Journal of neuroengineering and rehabilitation, 17, 1-21.
Koeppe, A., Bamer, F., & Markert, B. (2019). An efficient Monte Carlo strategy for elasto-plastic structures based on recurrent neural networks. Acta Mechanica, 230, 3279-3293.
Lee, J., Song, Y., & Shin, C. S. (2018). Effect of the sagittal ankle angle at initial contact on energy dissipation in the lower extremity joints during a single-leg landing. Gait & Posture, 62, 99-104. doi:https://doi.org/10.1016/j.gaitpost.2018.03.019
Logar, G., & Munih, M. (2015). Estimation of Joint Forces and Moments for the In-Run and Take-Off in Ski Jumping Based on Measurements with Wearable Inertial Sensors. Sensors, 15(5), 11258-11276. doi:10.3390/s150511258
Lu, Y., Wang, H., Hu, F., Zhou, B., & Xi, H. (2021). Effective recognition of human lower limb jump locomotion phases based on multi-sensor information fusion and machine learning. Med Biol Eng Comput, 59(4), 883-899. doi:10.1007/s11517-021-02335-9
Morgan, A. M., & O'Connor, K. M. (2019). Evaluation of an accelerometer to assess knee mechanics during a drop landing. J Biomech, 86, 125-131. doi:10.1016/j.jbiomech.2019.01.055
Mundt, M., Johnson, W. R., Potthast, W., Markert, B., Mian, A., & Alderson, J. (2021). A Comparison of Three Neural Network Approaches for Estimating Joint Angles and Moments from Inertial Measurement Units. Sensors, 21(13), 4535. Retrieved from https://www.mdpi.com/1424-8220/21/13/4535
Mundt, M., Koeppe, A., David, S., Witter, T., Bamer, F., Potthast, W., & Markert, B. (2020). Estimation of gait mechanics based on simulated and measured IMU data using an artificial neural network. Frontiers in bioengineering and biotechnology, 8, 41.
Mundt, M., Thomsen, W., Witter, T., Koeppe, A., David, S., Bamer, F., . . . Markert, B. (2020). Prediction of lower limb joint angles and moments during gait using artificial neural networks. Medical & biological engineering & computing, 58, 211-225.
Olson, M., Wyner, A., & Berk, R. (2018). Modern Neural Networks Generalize on Small Data Sets. https://proceedings.neurips.cc/paper_files/paper/2018/file/fface8385abbf94b4593a0ed53a0c70f-Paper.pdf
Palmerini, L., Klenk, J., Becker, C., & Chiari, L. (2020). Accelerometer-Based Fall Detection Using Machine Learning: Training and Testing on Real-World Falls. Sensors, 20(22). doi:10.3390/s20226479
Pichardo, A. W., Oliver, J. L., Harrison, C. B., Maulder, P. S., Lloyd, R. S., & Kandoi, R. (2019). Effects of Combined Resistance Training and Weightlifting on Motor Skill Performance of Adolescent Male Athletes. The Journal of Strength & Conditioning Research, 33(12), 3226-3235. doi:10.1519/jsc.0000000000003108
Prasanth, H., Caban, M., Keller, U., Courtine, G., Ijspeert, A., Vallery, H., & von Zitzewitz, J. (2021). Wearable Sensor-Based Real-Time Gait Detection: A Systematic Review. Sensors (Basel), 21(8). doi:10.3390/s21082727
Rapp, E., Shin, S., Thomsen, W., Ferber, R., & Halilaj, E. (2021). Estimation of kinematics from inertial measurement units using a combined deep learning and optimization framework. Journal of biomechanics, 116, 110229. doi:https://doi.org/10.1016/j.jbiomech.2021.110229
Seel, T., Raisch, J., & Schauer, T. (2014). IMU-Based Joint Angle Measurement for Gait Analysis. Sensors, 14(4), 6891-6909. doi:10.3390/s140406891
Stetter, B. J., Ringhof, S., Krafft, F. C., Sell, S., & Stein, T. (2019). Estimation of Knee Joint Forces in Sport Movements Using Wearable Sensors and Machine Learning. Sensors, 19(17), 3690. Retrieved from https://www.mdpi.com/1424-8220/19/17/3690
Stone, M. H., Fry, A. C., Ritchie, M., Stoessel-Ross, L., & Marsit, J. L. (1994). Injury potential and safety aspects of weightlifting movements. Strength & Conditioning Journal, 16(3), 15-21.
Suchomel, T. J., Comfort, P., & Stone, M. H. (2015). Weightlifting Pulling Derivatives: Rationale for Implementation and Application. Sports Medicine, 45(6), 823-839. doi:10.1007/s40279-015-0314-y
Suchomel, T. J., DeWeese, B. H., Beckham, G. K., Serrano, A. J., & Sole, C. J. (2014). The Jump Shrug: A Progressive Exercise Into Weightlifting Derivatives. Strength & Conditioning Journal, 36(3), 43-47. doi:10.1519/ssc.0000000000000064
Suchomel, T. J., Wright, G. A., & Lottig, J. (2014). Lower extremity joint velocity comparisons during the hang power clean and jump shrug at various loads. Paper presented at the ISBS-Conference Proceedings Archive.
Takei, S., Hirayama, K., & Okada, J. (2021). Comparison of the Power Output Between the Hang Power Clean and Hang High Pull Across a Wide Range of Loads in Weightlifters. The Journal of Strength & Conditioning Research, 35, S84-S88. doi:10.1519/jsc.0000000000003569
Tan, J.-S., Tippaya, S., Binnie, T., Davey, P., Napier, K., Caneiro, J. P., . . . Campbell, A. (2022). Predicting Knee Joint Kinematics from Wearable Sensor Data in People with Knee Osteoarthritis and Clinical Considerations for Future Machine Learning Models. Sensors, 22(2). doi:10.3390/s22020446
Varrecchia, T., De Marchis, C., Draicchio, F., Schmid, M., Conforto, S., & Ranavolo, A. (2020). Lifting Activity Assessment Using Kinematic Features and Neural Networks. Applied Sciences-Basel, 10(6). doi:10.3390/app10061989
Verhoeff, W. J., Millar, S. K., Oldham, A. R. H., & Cronin, J. (2020). Coaching the Power Clean: A Constraints-Led Approach. Strength & Conditioning Journal, 42(2), 16-25. doi:10.1519/ssc.0000000000000508
Vohlakari, K. (2022). Using Neural Networks to Estimate Lower Body Kinematics from IMU Data in Javelin Throwing.
Wilson, D. R., & Martinez, T. R. (2001). The need for small learning rates on large problems.
Wong, T.-T. (2015). Performance evaluation of classification algorithms by k-fold and leave-one-out cross validation. Pattern Recognition, 48(9), 2839-2846. doi:https://doi.org/10.1016/j.patcog.2015.03.009
Yoon, J. G. (2013). The Correlation between the Muscle Activity and Joint Angle of the Lower Extremity According to the Changes in Stance Width during a Lifting Task. J Phys Ther Sci, 25(8), 1023-1025. doi:10.1589/jpts.25.1023
Yu, B., Lin, C.-F., & Garrett, W. E. (2006). Lower extremity biomechanics during the landing of a stop-jump task. Clinical Biomechanics, 21(3), 297-305. doi:https://doi.org/10.1016/j.clinbiomech.2005.11.003
Zhou, Y. H., Rehman, R. Z. U., Hansen, C., Maetzler, W., Del Din, S., Rochester, L., . . . Lamoth, C. J. C. (2020). Classification of Neurological Patients to Identify Fallers Based on Spatial-Temporal Gait Characteristics Measured by a Wearable Device. Sensors, 20(15). doi:10.3390/s20154098