Abstract：Predicting walking speed is an effective way to improve the functionality of a feedback-controlled treadmill. A walking speed estimation method is proposed as a solution to this problem based on swing foot parameters using the empirical mode decomposition (EMD) method. The relationship between the user's average swing foot velocity and real walking speed is declared by theoretical derivation. To solve the problem of integral drift and accumulated error during acceleration double integration, the EMD method is used to decompose the acceleration signal into signals of different frequencies. After removing the drifting component, the velocity signal is reconstructed from signals whose frequencies are close to the intrinsic gait frequency. For the modal aliasing phenomenon caused by integral accumulation error, the ensemble EMD method is used to reconstruct the swing foot displacement signal. Based on the heel-strike and toe-off events, the period and foot displacements during the swing phase can be obtained. Thus, the average swing foot velocity is then calculated for walking speed estimation. The proposed method is tested under five different walking speed conditions and compared with the traditional double-integral method. Experimental results prove the effectiveness and benefit of the proposed method.
[1] SOUMAN J L, ROBUFFO P, GIORDANO, et al. Cyberwalk:Enabling unconstrained omnidirectional walking through virtual environments[J]. ACM Transactions on Applied Perception, 2011, 8(4):25. [2] FEASEL J, WHITTON M C, KASSLER L, et al. The integrated virtual environment rehabilitation treadmill system[J]. IEEE Transactions Neural System & Rehabilitation Engineering, 2011, 19(3):290-297. [3] CHRISTENSEN R R, HOLLERBACH J M, XU Y, et al. Inertial-force feedback for the tread-port locomotion interface[J]. Presence, 2014, 9(1):1-14. [4] FOSTY B, BEN-SADOUN G, SACCO G, et al. Accuracy and reliability of the RGB-D camera for measuring walking speed on a treadmill[J]. Gait & Posture, 2016, 48:113-119. [5] XU X, MCGORRY R W, CHOU L S, et al. Accuracy of the Microsoft KinectTM for measuring gait parameters during treadmill walking[J]. Gait & Posture, 2015, 34(2):1-7. [6] URSKA P, BRITTANY H, JUDITH E D. Validity and reliability of the Kinect for assessment of standardized transitional movements and balance:Systematic review and translation into practice[J]. Physical Medicine & Rehabilitation Clinics of North America, 2019, 30:399-422. [7] RAY N T, KNARR B A, HIGGINSON J S. Walking speed changes in response to novel user-driven treadmill control[J]. Journal of Biomechanics, 2018, 78:143-149. [8] CHA M, HAN S, KIM H, et al. User-driven treadmill using walking speed estimated from plantar pressure sensor[J]. Electronics Letters, 2017, 53(8):524-526. [9] KEIJSERS N L W, STOLWIJK N M, RENZENBRINK G J, et al. Prediction of walking speed using single stance force or pressure measurements in healthy subjects[J]. Gait & Posture, 2016, 43:93-95. [10] FUCHIOKA S, IWATA A, HIGUCHI Y, et al. The forward velocity of the center of pressure in the midfoot is a major predictor of gait speed in older adults[J]. International Journal of Gerontology, 2015, 9(2):119-122. [11] JOO S B, OH S E, SIM T, et al. Prediction of gait speed from plantar pressure using artificial neural networks[J]. Expert Systems with Applications, 2014, 41(16):7398-7405. [12] WANG K L, XU J. A speed regression using acceleration data in a deep convolutional neural network[J]. IEEE Access, 2019, 7:9351-9356. [13] MANNINI A, SABATINI A M. Walking speed estimation using foot-mounted inertial sensors:Comparing machine learning and strap-down integration methods[J]. Medical Engineering & Physics, 2014, 36(10):1312-1321. [14] LI Q, YOUNG M, NAING V, et al. Walking speed estimation using a shank-mounted inertial measurement unit[J]. Journal of Biomechanics, 2010, 43(8):1640-1643. [15] WANG L, SUN Y, LI Q, et al. Estimation of step length and gait asymmetry using wearable inertial sensors[J]. IEEE Sensors Journal, 2018, 18(9):3844-3851. [16] 吴哲明, 孙振国, 张文增, 等. 基于惯性测量单元旋转的陀螺漂移估计和补偿方法[J]. 清华大学学报(自然科学版), 2014, 54(9):1143-1147. WU Z M, SUN Z G, ZHANG W Z, et al. Gyroscope bias estimation and compensation by rotation of the inertial measurement unit[J]. Journal of Tsinghua University (Science and Technology), 2014, 54(9):1143-1147. (in Chinese). [17] 班朝, 任国营, 王斌锐, 等. 基于IMU的机器人姿态自适应EKF测量算法研究[J]. 仪器仪表学报, 2020, 41(2):33-39. BAN C, REN G Y, WANG B R, et al. Research on self-adaptive EKF algorithm for robot attitude measurement based on IMU[J]. Chinese Journal of Scientific Instrument, 2020, 41(2):33-39. (in Chinese). [18] SESSA S, ZECCA M, LIN Z, et al. A methodology for the performance evaluation of inertial measurement units[J]. Journal of Intelligent and Robotic Systems, 2013, 71:143-157. [19] ANWARY A R, YU H, VASSALLO M. Optimal foot location for placing wearable IMU sensors and automatic feature extraction for gait analysis[J]. IEEE Sensors Journal, 2018, 18(6):2555-2567. [20] BRZOSTOWSKI K. Toward the unaided estimation of human walking speed based on sparse modeling[J]. IEEE Transactions on Instrumentation and Measurement, 2018, 67(6):1389-1398. [21] ZHANG R, YANG H, HFLINGER F, et al. Adaptive zero velocity update based on velocity classification for pedestrian tracking[J]. IEEE Sensors Journal, 2017, 17(7):2137-2145. [22] YOON J, PARK H S, DAMIANO D. A novel walking speed estimation scheme and its application to treadmill control for gait rehabilitation[J]. Journal of NeuroEngineering and Rehabilitation, 2012, 9:62. [23] HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis[J]. Proceedings of the Royal Society A:Mathematical, Physical and Engineering Sciences, 1998, 454:903-995. [24] WU Z H, HUANG N E. Ensemble empirical mode decomposition:A noise-assisted data analysis method[J]. Advances in Adaptive Data Analysis, 2009, 1(1):1-41.
2021, Vol. 61 
