期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2015
卷号：3
期号：6
DOI：10.15680/ijircce.2015.0306014
出版社：S&S Publications
摘要：Highly developed artificial limb prostheses able of actuating many degrees of freedom (DOF) at thepresent open access obtainable. Pattern identification based algorithms with the purpose of make use of surfaceelectromyography (EMG) signals calculated beginning residual muscles demonstrate huge assure as multi-DOFcontrollers. Unfortunately, existing pattern recognition scheme is restricted to sequential manage of every DOF. Theprediction of instantaneous limb movement is an extremely attractive feature designed to manage of synthetic limbs. Inthis work, we proposed novel Hybrid Extreme Learning Machine (HELM) classification methods for the prediction ofthe limb movement and control them for individual through myoelectric signals. The HELM pattern recognitionmethods create a hybrid kernel function through fully mingle local kernel function forecast of the limb group. HELMpattern recognition suggests with the purpose of whichever classifier be able to be potentially working in thecalculation of instantaneous actions if prearranged in a distributed topology. In another way the proposed patternrecognition classifiers essentially able of simultaneous predictions, such as the HELM, were found to be presenttechnique more cost efficient, as they are able to be successfully working in their simplest form. The high accuracy ofthe HELM method suggests with the intention of pattern recognition techniques is able to be extensive to allowsimultaneous control, life-like actions, finally increasing their feature of life.
