基于肌电信号的人体下肢运动信息获取技术研究

《基于肌电信号的人体下肢运动信息获取技术研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、浙江大学计算机科学与技术学院博士学位论文基于肌电信号的人体下肢运动信息获取技术研究姓名：吴剑锋申请学位级别：博士专业：数字化艺术与设计指导教师：孙守迁20080701浙江大学博士学位论文摘要5．实现一个基于肌电信号的运动信息检测与反馈原型系统，对自主开发的老年人起立辅助座椅进行功能分析和评价，并验证本文所提出的理论、方法和技术的正确性与可行性．关键词：运动信息；肌电信号；特征提取；模式识别；肌肉力；关节力矩；智能辅助lV浙江大学博士学位论文Ab!她ctAbstractThehumanbodymayb

2、elostmovementabilitybecauseofVarioussubjectiVeandobjectivelimits．1tnotonlyafl’cctsthepatients’psychologicalandphysicalstatebutalsobring_ScompIexsocialprobIems．Howtohelpth锄liVeindependentlyhaSbecomeVe口import卸tnowadays．Theconceptofintclligentlimb—assiSte

3、dSyStemaimsatenhancinghumanmovementcapabiIities柚drctainingthefle灿biIityand佗elingfordirectope阳tionatthes锄etime．Sohowtoobtainthelimb’sinfo册“one懋ectiVelyandpreciSelyinreal-timebecomesurgentatpresent．EMGsignalsc柚direCtlyreflecthum卸muscles’内nction，basedonth

4、atthedissertationmainlyconcen臼锄es伽twoimportantcontents．oneishowtocalculatemusclefo佗eandjointtorques，andtheotherishowt0佗cognizeandpredictthehumanmotions．7rhemaincontentsofthisdisse似ionareshownasf0¨owing：1．Summarizedthetendencyandmethods内racquiringmotion

5、in内mationathomeandabroad．TheadVantages拥ddisadVantageswerepointed0ut．Na门mtedthemainresearchcontentsandresearchmethodsf-ortllisthesis．2．111erelationshipsbe铆e明di虢rentlower-limbmotionmodesandEMGsi印alswere卸alyzedthroughexpe—mentS．standingup，goingupanddownst

6、airsandwalkingwereperf．0mledastlledaily10、Ver-limbmotionintheexperiment．Theangleofeachjoint、Ⅳasmeasuredbymotioncapturcsysteln觚dtheactiV时leVeIsofmuscleswe坞accessedbythemuScleEMGsi印als．neresultsshowedthatdi腩rentlower-Jimbmotionshaddi仃brentmuscleexcitatio

7、ntimeandexcitationgrade．1tproVidedapracticalbasisforthef0¨owingstudy．3．111erecognitiontechnologyoflowe卜limbmotionmodeswas咖diedbasedonMulti．ClaSSSⅧ．FirStofall，theaims柏ddi衔cultiesofmodemcognition0nlower-limbmotionbased0nEMGsignalswereexpounded．Secondly，t

8、he锄alysisonEMGcharacteristicse加隐ctionhaSbeendone锄d锄eigenVe咖rspaceofmoderecognitionwaSbuilt．onthe0therhand，deVelopedanimpmVcdalgorithmbasedonMult．一ClaSsSⅥⅥ．neresultSofmoderecognitionexperimentsShowedthismethodcoulde行-ectivelyrcS0lvethelo