全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

利用平均影响值和概率神经网络的步态识别

DOI: 10.3969/j.issn.1006-7043.201309042

Keywords: 智能假肢, 步态识别, 平均影响值, 概率神经网络, 特征筛选

Full-Text   Cite this paper   Add to My Lib

Abstract:

为了实现对智能下肢假肢进行有效控制, 下肢步态(包括平地行走、上下楼梯和上下坡等)的有效识别是关键。先从提取不同步态下的特征值入手, 利用平均影响值(MIV)来实现变量的筛选, 并针对膝上截肢者的特点确定了6个特征值, 分别为髋关节角度的最大值、支撑前期均值、支撑中期均值、支撑中期标准差、摆动期标准差(即Mh、ISh、MSh、SWh、MSV、SWV), 然后利用概率神经网络(PNN)对本实验系统的5种步态进行准确识别, 并与BP神经网络(BPNN)识别步态进行比较, 试验结果表明将特征值用平均影响值方法筛选后, 用概率神经网络进行步态识别, 具有较好的识别率和识别速度, 其识别率与BP神经网络相比提高了10%以上, 验证了该方法的有效性和可行性。

References

[1]  PROTOPAPADAKI A, DRECHSLER W I. Hip, knee, ankle kinematics and kinetics during stair ascent and descent in healthy young individuals[J].Clinical Biomechanics, 2007, 22: 203-210.
[2]  张今瑜,王岚,张立勋.基于多传感器的实时步态检测研究[J]. 哈尔滨工程大学学报, 2007, 28(2): 218 -221.ZHANG Jinyu, WANG Lan, ZHANG Lixun. Real-time gait detection based on multiple sensors[J]. Journal of Harbin Engineering University, 2007, 28 (2): 218-221.
[3]  MATOVSKI D S, NIXON M S, MAHMOODI S, et al. The effect of time on gait recognition performance[J]. Information Forensics and Security, 2012, 7(2): 543-552.
[4]  ZENG Wei, WANG Cong. Human gait recognition via deterministic learning[J]. Neural Network, 2012, 35: 92-102.
[5]  CHEN Lingling, YANG Peng, XU Xiaoyun, et al. Above-knee prosthesis control based on posture recognition by support vector machine[C]//2008 IEEE International Conference on Robotics, Automation and Mechatronics. Chengdu, China, 2008: 307-312.
[6]  高云园,佘青山,孟明, 等.基于多源信息融合的膝上假肢步态识别方法[J]. 仪器仪表学报, 2010, 31(12): 2683-2688.GAO Yunyuan, SHE Qingshan, MENG Ming, et al. Recognition method based on multi- information fusion for gait patterns of above-knee prosthesis[J].Chinese Journal of Scientific Instrument, 2010,31 (12) : 2683-2688.
[7]  耿艳利,杨鹏,刘作军,等.大腿截肢者步速识别系统设计[J]. 中国康复医学杂志, 2012, 27(7): 651-653.GENG Yanli, YANG Peng, LIU Zuojun, et al. Pace recognition system for thigh amputees[J]. Chinese Journal of Rehabilitation Medicine, 2012, 27 (7): 651-653.
[8]  SAEID F, HADIS A, SHEIKH S M. Multi-view neural network based gait recognition[C]//World Academy of Science, Engineering and Technology. Istanbul, Turkey, 2010: 705-709.
[9]  NEDA K, SAEID R. Gait recognition for human identification using ensemble of LVQ neural networks[C]//2012 International Conference on Biomedical Engineering. Penang, Malaysia, 2012: 180-185.
[10]  SU F C, WU W L. Design and testing of a genetic algorithm neural network in the assessment of gait patterns[J]. Medical Engineering and Physics, 2000, 22(1): 67-74.
[11]  赵永伟,张二虎,鲁继文,等.多特征和多视角信息融合的步态识别[J]. 中国图象图形学报, 2009, 14(3): 388-393.ZHAO Yongwei, ZHANG Erhu, LU Jiwen, et al. Information fusion of multi-features and multi-angle for gait recognition[J]. Chinese Journal of Image and Graphics, 2009, 14 (3): 388-393.
[12]  XIAO Degui, YANG Lei. Gait recognition using Zernike moments and BP neural network[C]//2008 IEEE International Conference on Networking, Sensing and Control. Sanya, China, 2008: 418-423.
[13]  TEE Connie, KAH O G M, BENG J T A. Personal recognition using multi-angles gait sequences[J]. Communications in Computer and Information Science, 2011, 189(2): 497-508.
[14]  LEE Heesung, HONG Sungjun, KIM Euntai. Neural network ensemble with probabilistic fusion and its application to gait recognition[J]. Neurocomputing, 2009, 72(7-9): 1557-1564.
[15]  WANG Fei, WEN Shiguang, WU Chengdong. Gait pattern estimation for intelligent bionic leg[C]//Chinese Control and Decision Conference 2008. Yantai, China, 2008: 2216-2221.
[16]  张葛祥,金炜东,胡来招.基于量子遗传算法的特征选择算法[J]. 控制理论与应用, 2005, 22(5): 810-813.ZHANG Gexiang, JIN Weidong, HU Laizhao. Feature selection algorithm based on quantum genetic algorithm[J].Control Theory and Applications, 2005, 22(5): 810-813.
[17]  FU Zhongguang, QI Minfang,JING Yuan. Regression forecast of main steam flow based on mean impact value and support vector regression[C]//Power and Energy Engineering Conference (APPEEC). Shanghai, China, 2012: 27-29.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133