Real-time prediction method for lower limb movement joint angle based on similar rule search

Abstract

The invention discloses a real-time prediction method for a lower limb motion joint angle based on similar rule search. The real-time prediction method comprises the steps that a lower limb motion rule is obtained; for a single joint, six-axis data and a joint angle angle in periodic movement of the lower limb are obtained, and a motion data column vector and a joint angle label vector are obtained at each sampling moment; motion data frames and joint angle label frames are continuously obtained through a sliding window, and a joint motion rule database is constructed; the current motion frameobtained by the real-time sliding window is matched with a historical frame sequence in a joint motion rule database, the historical frame with the highest matching degree is found out, and the corresponding sampling point coordinate is taken as the optimal matching position; and the value of the joint angle label frame after the sampling point is taken as a joint angle prediction value of the next sampling point moment to realize joint angle prediction of a plurality of continuous moments in the future. According to the real-time prediction method, the hysteresis quality of the current wearable robot following control can be effectively improved, and the wearing comfort is improved.

Description

technical field [0001] The invention relates to the technical field of real-time continuous prediction of human motion intentions, in particular to a method for real-time prediction of lower limb motion joint angles based on similarity rule search. Background technique [0002] With the continuous development of wearable robot technology, its control requirements are gradually changing and advancing from passive control to active control. Among its key technologies, the control of compliance has become a bottleneck restricting its development, and the perception and prediction of human motion intention is the key to restricting compliance control. Currently, human intention perception prediction methods mainly include intention perception prediction based on EEG, EMG, and physical motion sensors. The EEG signal is generated at the front end of the intention transmission process, ahead of the movement of the lower limbs, but it is difficult to identify and cannot be applied ...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem of real-time prediction of the joint angle of the lower limbs, realize the real-time construction of the prediction model and the real-time accurate prediction of the joint angle, and improve the compliance control performance of the existing wearable robot. The real-time prediction method of the joint angle of the lower limbs uses physical motion sensors to obtain the movement laws of the lower limbs, builds a database of joint movement laws, searches based on similar laws and uses historical laws to realize future movement predictions, so as to realize real-time prediction of the joint angles of the lower limbs. The invention not only has a fast prediction speed, but also realizes the online real-time construction of the prediction model, and the algorithm has high robustness

Images