Dynamic Stability Analysis And Gait Tracking Control For The Lower Extremity Exoskeleton

The lower extremity exoskeleton is a typical mechatronic system which is wore by the walker to enhance his strength and endurance. It has a broad application prospect in fields of rehabilitation, military and transportation since it achieves the complementary advantages of human and robot. The lower extremity exoskeleton tracks the wearer’s biped walking, which has good walking flexibility but poor dynamic stability. In addition, due to the time consuming of the sensing system, drive system and control system, the lower extremity exoskeleton is out of step with the wearer because of time delay. This paper mainly discusses these two problems and the main work is shown as follows:(1) The physiological structure and motion mechanism of the human body are studied to design the mechanical structure of the lower extremity exoskeleton. The dynamic model is built by using the Lagrange equation. Then the drive system of the lower extremity exoskeleton is designed according to the calculation results of the dynamics model.(2) For overcoming the shortage of the zero movement point(ZMP) method, an improved ZMP dynamic stability evaluation method is proposed to analyze the dynamic stability of the lower extremity exoskeleton. Next we study the dynamic stability of the lower extremity exoskeleton in the coronal and sagittal plane by respectively using the ZMP stability criterion and the improved ZMP dynamic stability evaluation method. Then the methods to improve the dynamic stability in the coronal and sagittal plane are obtained.(3) For the time delay of the lower extremity exoskeleton tracking the wearer, we make the lower extremity exoskeleton to track the predictive gait of the wearer. Some gait prediction methods are studied to select the optimal prediction method, such as the gait prediction method based on the characteristic analysis of human, the gait prediction method based on the Kalman filter, the gait prediction method based on the Newton predictor and its improved form. The gait prediction method based on the improved Newton predictor is selected to eliminate the influence of the time delay. Then we reduce the prediction time to suppress the effect of vibration on gait prediction when the lower extremity exoskeleton vibrates. (4) The Gait Tracking Control strategy is proposed. Then we build the experimental platform of the lower extremity exoskeleton on which we do the single joint control experiment, gait predictive experiment and gait predictive following control experiments. The experimental results show that the gait prediction method based on the improved Newton predictor can suppress the effect of the time delay very well and the Gait Tracking Control can improve the dynamic stability of the lower extremity exoskeleton.