摘要
在研究了人体运动的基础上提出了一种基于改进的Elman网络模型的人体肌肉动力学模型,给出了网络的学习算法,并以运动员举重提铃动作的下肢肌肉运动为研究对象,建立了最优关节力矩逼近的Elman肌肉动力学网络模型.结果表明该模型通过预测肌肉神经控制激活参数,较好地拟合了关节力矩曲线.
In this paper human motor control is studied and a new dynamical model of human muscle following joint moment based on Elman neural work is proposed. Then a BPTT learning algorithm is discussed. Finally a Ekman network is built up to model athletic motion of lifting up barbell about weight-lifting, the result shows that this model predicts muscle activation very well, and it has good prospect as well.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2004年第3期374-379,共6页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金(No.60343006
60375027)
安徽省自然科学基金(No.03042304)