Identification of EMG-force system using the second-order Volterra model

Conventional method to demodulate force signal from corresponding electromyographic (EMG) signal is to low-pass filter the rectified EMG signal. This method gives rise to a poor performance when voluntary muscle contraction level increases or decreases rapidly. In this work, the second-order Volterra model was adopted to represent the EMG/force system and the nonlinear system identification method was used to obtain the system parameters or kernel. The EMG and muscle force data were obtained from human biceps during isometric voluntary contraction. The results of this work show that the second-order Volterra model provides a faster response in the transient phase and smaller error in the steady state than conventional linear low-pass filter models