Lower Extreme Carrying Exoskeleton Robot Adative Control Using Wavelet Neural Networks

Abstract

Using the wavelet neural networks, an adaptive control system, with two wavelet neural networks as controller and dynamics model identifier respectively, is developed for lower extreme carrying exoskeleton robot. Because the wavelet neural networks have the ability to approximate nonlinear functions and good advantage of time-frequency localization properties, this system can identify nonlinear system dynamic characters more precisely, and can map more complex control strategies. Results show that this control system is more effective than those based on normal controller, where the exoskeleton tracking precision is high and the operator feels very little torque.