
浏览全部资源
扫码关注微信
福建师范大学体育科学学院(福州 350117)
LIN Yiwei (E-mail: 116132019022@student.fjnu.edu.cn)
Received:31 December 2024,
Revised:2025-04-08,
Published:20 April 2026
移动端阅览
林奕炜,张涵宇.基于神经网络模型预测跑步中地面反作用力研究进展[J].北京生物医学工程,2026,45(2):207-214.
LIN Yiwei,ZHANG Hanyu.Research progress on predicting ground reaction force in running based on neural network models[J].Beijing Biomedical Engineering,2026,45(2):207-214.
林奕炜,张涵宇.基于神经网络模型预测跑步中地面反作用力研究进展[J].北京生物医学工程,2026,45(2):207-214. DOI: 10.3969/j.issn.1002-3208.2026.02.014.
LIN Yiwei,ZHANG Hanyu.Research progress on predicting ground reaction force in running based on neural network models[J].Beijing Biomedical Engineering,2026,45(2):207-214. DOI: 10.3969/j.issn.1002-3208.2026.02.014.
在跑步运动中,量化地面反作用力(ground reaction force
GRF)对于优化跑步经济性、提高运动表现以及损伤预防具有重要意义。然而,传统的GRF测量方法高度依赖复杂且昂贵的实验室设备,限制了其广泛应用。近年来基于神经网络模型拟合人体运动学和动力学的方法为简化GRF测量提供了新的解决方案。本文就国内外关于使用神经网络模型预测GRF的研究进行系统回顾。可穿戴设备采集的下肢运动学数据能够为基于神经网络模型预测GRF提供足够的信息;在模型选择上,多数研究倾向于使用长短期记忆网络(long short-term memory network
LSTM)或者能提供动态记忆的网络模型,以处理GRF的时序信号特性并实现精准预测。基于神经网络模型的预测方法配合可穿戴设备采集的数据不仅有效摆脱传统测力台的依赖,还在预测GRF方面表现出了较高的准确性,为跑步损伤预防和训练优化提供了实用的参考依据。然而,神经网络模型对大规模高质量数据的依赖性较高,模型的泛化能力在多样化运动场景中的表现尚不充分,此外,计算成本较高和实时性能有限的问题也限制了其在部分场景中的应用。因此,未来的研究需要进一步探索轻量化模型设计、跨场景适配能力的提升以及实时预测技术的发展,以推动神经网络模型在运动监测中的实际应用与普及。
In the sport of running, quantifying the ground reaction force (GRF) is of great significance for optimizing running economy, enhancing athletic performance, and preventing injuries. However, traditional GRF measurement methods are highly dependent on complex and expensive laboratory equipment, which limits their wide application. In recent years, methods based on neural network modeling to fit human kinematics and dynamics have provided new solutions to simplify GRF measurements. This paper presents a systematic review of domestic and international studies on the prediction of GRF by using neural network models. Lower limb kinematic data collected by wearable devices can provide sufficient information for GRF prediction based on neural network models; in terms of model selection, most studies tend to use long short-term memory network (LSTM) or network models that can provide dynamic memory to deal with the temporal signal characteristics of GRF and achieve accurate prediction. The neural network model-based prediction method with the data collected by wearable devices not only effectively eliminates the reliance on traditional ergometers, but also shows high accuracy in predicting GRF, which provides a practical reference for running injury prevention and training optimization. However, the neural network model has a high dependence on large-scale high-quality data, and the generalization ability of the model is not yet sufficient in diverse sports scenarios, in addition, the problems of high computational cost and limited real-time performance also limit its application in some scenarios. Therefore, future research needs to further explore the lightweight model design, the improvement of cross-scene adaptation ability, and the development of real-time prediction techniques to promote the practical application and popularization of neural network models in motion monitoring.
Lee DC, Pate RR, Lavie CJ, et al. Leisure-time running reduces all-cause and cardiovascular mortality risk[J]. Journal of the American College of Cardiology, 2014, 64(5): 472-481.
Menheere D, Janssen M, Funk M, et al. Runner’s perceptions of reasons to quit running: influence of gender, age and running-related characteristics[J]. International Journal of Environmental Research and Public Health, 2020, 17(17): 6046.
Johnson CD, Tenforde AS, Outerleys J, et al. Impact-related ground reaction forces are more strongly associated with some running injuries than others[J]. The American Journal of Sports Medicine, 2020, 48(12): 3072-3080.
Riley PO, Dicharry J, Franz J, et al. A kinematics and kinetic comparison of overground and treadmill running[J]. Medicine Science in Sports Exercise, 2008, 40(6): 1093-1100.
Kriegeskorte N, Golan T. Neural network models and deep learning[J]. Current Biology, 2019, 29(7): R231-R236.
Frederick EC, Hagy JL. Factors affecting peak vertical ground reaction forces in running[J]. International Journal of Sport Biomechanics, 1986, 2(1): 41-49.
Clark KP, Ryan LJ, Weyand PG. A general relationship links gait mechanics and running ground reaction forces[J]. Journal of Experimental Biology, 2017, 220(Pt 2): 247-258.
Scheltinga BL, Kok JN, Buurke JH, et al. Estimating 3D ground reaction forces in running using three inertial measurement units[J]. Frontiers in Sports and Active Living, 2023, 5: 1176466.
Wang D, Li S, Song Q, et al. Predicting vertical ground reaction force in rearfoot running: a wavelet neural network model and factor loading[J]. Journal of Sports Sciences, 2023, 41(10): 955-963.
Honert EC, Hoitz F, Blades S, et al. Estimating running ground reaction forces from plantar pressure during graded running[J]. Sensors, 2022, 22(9): 3338.
Ngoh KJH, Gouwanda D, Gopalai AA, et al. Estimation of vertical ground reaction force during running using neural network model and uniaxial accelerometer[J]. Journal of Biomechanics, 2018, 76: 269-273.
Pogson M, Verheul J, Robinson MA, et al. A neural network method to predict task-and step-specific ground reaction force magnitudes from trunk accelerations during running activities[J]. Medical Engineering Physics, 2020, 78: 82-89.
Komaris DS, Pérez-Valero E, Jordan L, et al. Predicting three-dimensional ground reaction forces in running by using artificial neural networks and lower body kinematics[J]. IEEE Access, 2019, 7: 156779-156786.
Niemelä M, Kulmala JP, Kauppi JP, et al. Prediction of active peak force using a multilayer perceptron[J]. Sports Engineering, 2017, 20: 213-219.
Donahue SR, Hahn ME. Estimation of gait events and kinetic waveforms with wearable sensors and machine learning when running in an unconstrained environment[J]. Scientific Reports, 2023, 13(1): 2339.
Sharma D, Davidson P, Müller P, et al. Indirect estimation of vertical ground reaction force from a body-mounted INS/GPS using machine learning[J]. Sensors, 2021, 21(4): 1553.
Johnson WR, Mian A, Robinson MA, et al. Multidimensional ground reaction forces and moments from wearable sensor accelerations via deep learning[J]. IEEE Transactions on Biomedical Engineering, 2020, 68(1): 289-297.
Hajizadeh M, Clouthier AL, Kendall M, et al. Predicting vertical and shear ground reaction forces during walking and jogging using wearable plantar pressure insoles[J]. Gait Posture, 2023, 104: 90-96.
Bach MM, Dominici N, Daffertshofer A. Predicting vertical ground reaction forces from 3D accelerometry using reservoir computers leads to accurate gait event detection[J]. Frontiers in Sports and Active Living, 2022, 4: 1037438.
Wouda FJ, Giuberti M, Bellusci G, et al. Estimation of vertical ground reaction forces and sagittal knee kinematics during running using three inertial sensors[J]. Frontiers in Physiology, 2018, 9: 218.
Alcantara RS, Edwards WB, Millet GY, et al. Predicting continuous ground reaction forces from accelerometers during uphill and downhill running: a recurrent neural network solution[J]. PeerJ, 2022, 10: e12752.
Nedergaard NJ, Verheul J, Drust B, et al. The feasibility of predicting ground reaction forces during running from a trunk accelerometry driven mass-spring-damper model[J]. PeerJ, 2018, 6: e6105.
Lieberman DE, Venkadesan M, Werbel WA, et al. Foot strike patterns and collision forces in habitually barefoot versus shod runners[J]. Nature, 2010, 463(7280): 531-535.
Gupta TK, Raza K. Optimization of ANN architecture: a review on nature-inspired techniques[J]. Machine Learning in Bio-Signal Analysis and Diagnostic Imaging, 2019: 159-182.
Sherstinsky A. Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network[J]. Physica D: Nonlinear Phenomena, 2020, 404: 132306.
Jaeger H, Haas H. Harnessing nonlinearity: predicting chaotic systems and saving energy in wireless communication[J]. Science, 2004, 304(5667): 78-80.
Scheltinga BL, Usta H, Reenalda J, et al. Estimating vertical ground reaction force during running with 3 inertial measurement units[J]. Journal of Biomedical Engineering and Biosciences (JBEB), 2022, 9(1): 31-38.
Guo T, Zhang T, Lim E, et al. A review of wavelet analysis and its applications: challenges and opportunities[J]. IEEE Access, 2022, 10: 58869-58903.
Molavian R, Fatahi A, Abbasi H, et al. Artificial intelligence approach in biomechanics of gait and sport: a systematic literature review[J]. Journal of Biomedical Physics Engineering, 2023, 13(5): 383-402.
Zamparo P, Perini R, Orizio C, et al. The energy cost of walking or running on sand[J]. European Journal of Applied Physiology and Occupational Physiology, 1992, 65(2): 183-187.
Voloshina AS, Ferris DP. Biomechanics and energetics of running on uneven terrain[J]. Journal of Experimental Biology, 2015, 218(Pt 5): 711-719.
Chan ZYS, Zhang JH, Au IPH, et al. Gait retraining for the reduction of injury occurrence in novice distance runners: 1-year follow-up of a randomized controlled trial[J]. The American Journal of Sports Medicine, 2018, 46(2): 388-395.
0
Views
0
下载量
0
CSCD
Publicity Resources
Related Articles
Related Author
Related Institution
京公网安备11010802024621