Lower Limb Locomotion Modes Recognition Based on Multiple-source Information and General Regression Neural Network
LIU Lei1, YANG Peng1,2, LIU Zuojun1,2
1. School of Control science and Engineering, Hebei University of Technology, Tianjin 300130, China;
2. Engineering Research Center of Intelligent Rehabilitation and Detecting Technology, Tianijin 300130, China
刘磊, 杨鹏, 刘作军. 基于多源信息和广义回归神经网络的下肢运动模式识别[J]. 机器人, 2015, 37(3): 310-317.DOI: 10.13973/j.cnki.robot.2015.0310.
LIU Lei, YANG Peng, LIU Zuojun. Lower Limb Locomotion Modes Recognition Based on Multiple-source Information and General Regression Neural Network. ROBOT, 2015, 37(3): 310-317. DOI: 10.13973/j.cnki.robot.2015.0310.
In order to improve recognition rate of lower limb locomotion modes, a method based on multiple-source information and general regression neural network (GRNN) is proposed. Users' daily lower limb locomotion modes are decomposed into different segments to form the recognition goals using the plantar pressure sensor. For surface electromyography (sEMG) signal, three features are used, i.e. skewness, kurtosis, and power spectral entropy. The hip joint angle is chosen as leg features. In order to decrease the time for training the models and to prevent overfitting, principal component analysis (PCA) is used to reduce the dimension of the extracted features. GRNN is used to recognize 3 kinds of motions, namely stairs ascent, stairs descent and level-ground walking. The experimental results show that the recognition correct rate is 90.16%.
[1] 喻洪流,张晓玉.假肢矫形器原理与应用[M].1版.南京:东南大学出版社,2011:2--3. ewline Yu H L, Zhang X Y. Principle and application of the prosthetic brace[M]. 1st ed. Nanjing: Southeast University Press, 2011: 2--3.[2] Huang H, Kuiken T A, Lipschutz R D. A strategy for identifying locomotion modes using surface electromyography[J]. IEEE Transactions on Biomedical Engineering, 2009, 56(1): 65--73.[3] 孟明,佘青山,罗志增.HMM在下肢表面肌电信号步态识别中的应用[J].华中科技大学学报:自然科学版,2011,39(S2):176--179.Meng M, She Q S, Luo Z Z. The application of HMM in gait recognition using lower limb sEMG[J]. Journal of Huazhong University of Science & Technology: Natural Science Edition, 2011, 39(S2): 176--179.[4] 刘秀云.基于EMG-KJA神经肌骨动力学模型的下肢动作模式识别及运动轨迹预测[D].天津:天津大学,2012. Liu X Y. Motion pattern recognition & kinematic trajectory prediction of lower limb based on EMG-KJA neuro-musculo-skeletal model[D]. Tianjin: Tianjin University, 2012.[5] 佘青山,高云园,孟明,等.下肢EMG的小波支持向量机多类识别方法[J].华中科技大学学报:自然科学版,2010,38(10):75--79. She Q S, Gao Y Y, Meng M, et al. Multiclass recognition of lower limb EMG using wavelet SVM[J]. Journal of Huazhong University of Science & Technology: Natural Science Edition, 2010, 38(10): 75--79.[6] 苟斌,刘作军,赵丽娜,等.基于相关性分析的下肢假肢步行模式预识别方法[J].东南大学学报:自然科学版,2013,43(1):192--196. Gou B, Liu Z J, Zhao L N, et al. Walking mode pre-judgment of lower limb prosthesis based on correlation analysis[J]. Journal of Southeast University: Natural Science Edition, 2013, 43(1): 192--196.[7] 赵丽娜,刘作军,苟斌,等.基于隐马尔可夫模型的动力型下肢假肢步态预识别[J].机器人,2014,36(3):337-341.Zhao L N, Liu Z J, Gou B, et al. Gait pre-recognition of dynamic lower limb prosthesis based on hidden Markov model[J]. Robot, 2014, 36(3): 337-341.[8] BenAbdelkader C, Cutler R, Davis L. Motion based recognition of people in eigengait space[C]//IEEE International Conference on Automatic Face and Gesture Recognition. Piscataway, USA: IEEE, 2002: 267-272.[9] 佟丽娜,侯增广,彭亮,等.\columnbreak基于多路sEMG时序分析的人体运动模式识别方法[J].自动化学报,2014,40(5):810-821. Tong L N, Hou Z G, Peng L, et al. Multi-channel sEMG time series analysis based human motion recognition method[J]. Acta Automatica Sinica, 2014, 40(5): 810-821.[10] 熊静.下肢姿态检测及运动状态预测算法研究[D].杭州:杭州电子科技大学,2009. Xiong J. Detection of lower limb posture and research on prediction algorithm for movement state[D]. Hangzhou: Hangzhou Dianzi University, 2009.[11] Tkach D, Huang H, Kuiken T A, et al. Study of stability of time-domain features for electromyographic pattern recognition[J]. Journal of Neuroengineering and Rehabilitation, 2010, 7: No.21.[12] Hargrove L J, Li G L, Englehart K B, et al. Principal components analysis preprocessing for improved classification accuracies in pattern-recognition-based myoelectric control[J]. IEEE Transactions on Biomedical Engineering, 2009, 56(5): 1407-1414.[13] Su I J,Tsai C C,Sung W T. Comparison of BP and GRNN\linebreak algorithm for factory monitoring[J]. Applied Mechanics and Materials, 2011, 52-54: 2105-2110.[14] 冯志鹏,宋希庚,薛冬新,等.基于广义回归神经网络的时间序列预测研究[J].振动、测试与诊断,2003,23(2):105-109. Feng Z P, Song X G, Xue X D, et al. General regression neural network based prediction of time series[J]. Journal of Vibration, Measurement & Diagnosis, 2003, 23(2): 105-109.[15] 吴剑锋,吴群,孙守迁.简约支持向量机分类算法在下肢动作识别中的应用研究[J].中国机械工程,2011,22(4):433-438. Wu J F, Wu Q, Sun S Q. Research on classification algorithm\linebreak of reduced support vector machine for low limb movement\linebreak recognition[J]. China Mechanical Engineering, 2011, 22(4): 433-438.[16] 刑秀玉,刘鸿宇,黄武.基于加速度的小波能量特征及样本熵组合的步态分类算[J].传感技术学报,2013,26(4):545-549.Xing X Y, Liu H Y, Huang W. Gait pattern classification with wavelet energy and sample entropy based on acceleration signals[J]. Chinese Journal of Sensors and Actuators, 2013, 26(4): 545-549.