Ankle intention detection algorithm using electromyography signal

Abstract In this study, an ankle intention detection algorithm was developed to calculate the torque user wants exert from starting user’s EMG signal. Since subtalar joint axis of is very important for stability, intent also calculates eversion motion axis. A dry sensor used measure signal, and biaxial measurement device manufactured perform experiment. The experiment conducted on four healthy subjects (mean ± SD: height, 177.6 7.3 cm; weight 70.2 8.9 kg; age, 27 2 years), signals were measured. Using experimental results a neural network, we algorithm. When you input estimates eversion, dorsiflexion, plantar flexion. error 0.37 Nm (subtalar) 0.57 (talocrural), which 0.5% 1.5% (talocrural) required walking. more accurate stable exoskeleton robot control becomes possible.