DocumentCode :
3660076
Title :
A human motion prediction algorithm for Non-binding Lower Extremity Exoskeleton
Author :
Min Wang;Xinyu Wu;Duxin Liu;Can Wang;Ting Zhang;Pingan Wang
Author_Institution :
Key Laboratory of Human-Machine Intelligence-Synergy Systems, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, China
fYear :
2015
Firstpage :
369
Lastpage :
374
Abstract :
This paper introduces a novel approach to predict human motion for the Non-binding Lower Extremity Exoskeleton (NBLEX). Most of the exoskeletons must be attached to the pilot, which exists potential security problems. In order to solve these problems, the NBLEX is studied and designed to free pilots from the exoskeletons. Rather than applying Electromyography (EMG) and Ground Reaction Force (GFR) signals to predict human motion in the binding exoskeleton, the non-binding exoskeleton robot collect the Inertial Measurement Unit (IMU) signals of the pilot. Seven basic motions are studied, each motion is divided into four phases except the standing-still motion which only has one motion phase. The human motion prediction algorithm adopts Support Vector Machine (SVM) to classify human motion phases and Hidden Markov Model (HMM) to predict human motion. The experimental data demonstrate the effectiveness of the proposed algorithm.
Keywords :
"Hidden Markov models","Prediction algorithms","Exoskeletons","Support vector machines","Accuracy","Classification algorithms","Training"
Publisher :
ieee
Conference_Titel :
Information and Automation, 2015 IEEE International Conference on
Type :
conf
DOI :
10.1109/ICInfA.2015.7279315
Filename :
7279315
Link To Document :
بازگشت