仪器仪表学报
儀器儀錶學報
의기의표학보
CHINESE JOURNAL OF SCIENTIFIC INSTRUMENT
2014年
z2期
7-11
,共5页
叶斌%李玉榕%陈建国%吴福春%杜民
葉斌%李玉榕%陳建國%吳福春%杜民
협빈%리옥용%진건국%오복춘%두민
Android%KOA%扩展Kalman滤波%规则归纳法
Android%KOA%擴展Kalman濾波%規則歸納法
Android%KOA%확전Kalman려파%규칙귀납법
Android%KOA%extended Kalman filter%rule induction
本文设计了一种便捷式的膝骨性关节炎(KOA)运动监测系统,KOA患者可以通过该监测系统了解自身运动规范性并作出适当的调整。该监测系统将装有MMA7361加速度传感器的蓝牙模块穿戴在人体的下肢进行三轴加速度信号的采集,并将三轴加速度信号通过蓝牙发送至Android智能手机上。 Android手机通过扩展Kalman滤波算法将采集到的加速度信号进行滤波,并通过规则归纳法对动作规范性进行识别。经过大量实验测试结果统计,对保持时间太短、运动腿弯曲、抬腿速度太快这三类动作训练的平均识别率分别为89.6%,89.3%,89.5%。结果表明:基于Android智能手机平台构建的KOA运动监测系统具有便捷、成本低等优点,而且能够较好的识别患者运动中出现的不规范动作,满足运动监测的应用要求。
本文設計瞭一種便捷式的膝骨性關節炎(KOA)運動鑑測繫統,KOA患者可以通過該鑑測繫統瞭解自身運動規範性併作齣適噹的調整。該鑑測繫統將裝有MMA7361加速度傳感器的藍牙模塊穿戴在人體的下肢進行三軸加速度信號的採集,併將三軸加速度信號通過藍牙髮送至Android智能手機上。 Android手機通過擴展Kalman濾波算法將採集到的加速度信號進行濾波,併通過規則歸納法對動作規範性進行識彆。經過大量實驗測試結果統計,對保持時間太短、運動腿彎麯、抬腿速度太快這三類動作訓練的平均識彆率分彆為89.6%,89.3%,89.5%。結果錶明:基于Android智能手機平檯構建的KOA運動鑑測繫統具有便捷、成本低等優點,而且能夠較好的識彆患者運動中齣現的不規範動作,滿足運動鑑測的應用要求。
본문설계료일충편첩식적슬골성관절염(KOA)운동감측계통,KOA환자가이통과해감측계통료해자신운동규범성병작출괄당적조정。해감측계통장장유MMA7361가속도전감기적람아모괴천대재인체적하지진행삼축가속도신호적채집,병장삼축가속도신호통과람아발송지Android지능수궤상。 Android수궤통과확전Kalman려파산법장채집도적가속도신호진행려파,병통과규칙귀납법대동작규범성진행식별。경과대량실험측시결과통계,대보지시간태단、운동퇴만곡、태퇴속도태쾌저삼류동작훈련적평균식별솔분별위89.6%,89.3%,89.5%。결과표명:기우Android지능수궤평태구건적KOA운동감측계통구유편첩、성본저등우점,이차능구교호적식별환자운동중출현적불규범동작,만족운동감측적응용요구。
In this paper, a convenient type of knee osteoarthritis (KOA) movement monitoring system is designed so that KOA patients can make appropriate adjustments when their movements are not standard. In the monitoring system, a MMA7361 acceleration sensor together with Bluetooth module wears in the body of the lower limbs for acquiring the triaxial acceleration signal, which will be sent to the Android smart phone by the Bluetooth module. Then, the extended Kalman filter is used for filtering the collected acceleration signal, and also the rule induction is exploited to identify the standard action. After a large number of experimental, including hold time is too short, curved legs movement, leg movement too fast, the statistics for these three average recognition rates were 89.6%, 89.3%, 89.5%, respectively. The results show that KOA movement monitoring system based on Android smart phone is convenient, low cost, and can well recognize the non-standard movement of patients, which meet the requirement of the movement monitoring system.