Surface electromyography signal analysis and pattern recognition based on Hilbert-Huang transform

The surface electromyography signal of stump was applied to recognize the intention of disable people, and control the intelligent above-knee prosthesis. Since the electromyography signal was non-stationary and non-linear signal, Hilbert-Huang transform was applied to extract the frequency feature. The time and frequency features of surface electromyography signal recorded from biceps femoris of the disable people´s stump, were applied to recognize the conversion of movement pattern. The result of this study indicates that this method has preferable recognize effect, and can be applied to analyze the nonlinear relationship between different movement patterns and electromyography signal.