论文部分内容阅读
提出了一种肌肉等长收缩力估计与肌肉固有特性表征的新方法,称为能量核方法.此方法的初衷在于将表贴EMG(肌电图)信号转变为平面内的相图,并将相图上状态点的分布核心称作能量核,而噪声信号的分布核心称为噪声核.基于相图的统计特征,将一段EMG信号近似为简谐振子,简称EMG振子.本文建立了控制信号(EMG)与输出信号(力/功率)之间的关系,并提出用EMG的特征能量来表征肌肉力.另一方面,通过对能量核与噪声核的计算,能够得到噪声与EMG信号的自然频率并实现直观的信噪识别与分离.实验结果表明,特征能量对等长收缩力的表征度令人满意,并且由于结合了RMS与MPF两种方法的优点,此方法具有很高的鲁棒性;而特定肌肉的EMG自然频率不取决于MU放电频率,故其反映了肌肉的固有特性.此模型体现的物理意义为EMG信号的理解与分析提供了新的启发.
A new method for estimating muscle isometric contractility and characterization of muscular intrinsic properties is proposed, which is called energy nuclear method.The original intention of this method is to transform EMG (EMG) signal into in-plane phase diagram, The distribution of state points on the phase diagram is called the energy kernel, while the distribution of the noise signal is called the noise kernel. Based on the statistical characteristics of the phase diagram, a piece of EMG signal is approximated by a simple harmonic oscillator (EMG) (EMG) and the output signal (force / power), and proposes to use the EMG’s characteristic energy to characterize the muscle force.On the other hand, by calculating the energy kernel and noise kernel, we can get the natural noise and EMG signal Frequency and to achieve an intuitive recognition and separation of the signal and noise.The experimental results show that the characteristic energy has a satisfactory representation of the isometric contraction force and this method is robust due to the merits of both the RMS and MPF methods However, the natural frequency of EMG of a specific muscle does not depend on the frequency of MU discharge, so it reflects the inherent characteristics of muscle.The physical meaning embodied by this model provides a new inspiration for the understanding and analysis of EMG signals.