Exoskeleton movement collaboration method based on long and short term memory network

Abstract

The invention discloses an exoskeleton movement collaboration method based on a long and short term memory network model. The method comprises the following operation steps: a, constructing a power exoskeleton hardware system; b, constructing a separated joint sensing system; c, performing joint angle collaborative movement; d, predicting a long and short-term memory network gait track; and e, predicting gait movement collaboration. On the basis of the LSTM model method, real-time data of the power exoskeleton in the joint cooperative movement process is used for predicting the movement track, the method can be used for man-machine cooperative movement gait track planning of a lower limb exoskeleton robot, and the problem of movement lag caused by cooperative control using a mechanical sensor can be solved.

Description

technical field [0001] The invention relates to a long-short-term memory network-based exoskeleton motion coordination method, which is applied to the research field of power exoskeleton gait cooperative motion control. Background technique [0002] The powered exoskeleton is worn on the outside of the human body, and the sensor system detects the position and posture of the exoskeleton itself and the movement intention of the human body in real time, and realizes the coordinated movement with the human body through the drive system, assisting the movement of the human body, achieving the enhancement of human body strength or assisting The goal of human movement. Many studies regard gait trajectory as a time series of joint angle changes over time. In this way, the prediction of gait trajectory is essentially the prediction of time series, that is, the future value is the result of prediction based on previous observations. Gait trajectory prediction methods based on machin...

AI Technical Summary

Problems solved by technology

On the other hand, the progress made in the current gait trajectory prediction method mainly focuses on the prediction based on human motion trajectory rather than the prediction based on powered exoskeleton trajectory, although this is the exploration of human motion during the pre-development process of powered exoskeleton prototype. It is a necessary stage of the characteristics, but the gait trajectory prediction lacking the real power exoskeleton motion information is still only in the simulation stage in a realistic sense. If the gait trajectory can be predicted and added to the control algorithm, it will have a great impact on the control response time. The delay problem caused by adding feed-forward items to compensate or the system's own data operation is helpful

Images