Analysis Of The Motion State Of The Humanoid Elbow Joint Based On Muscle's Drive Characteristics

The realization of man-machine collaboration makes the interactive activities between man-machines more natural and supple,which is an important direction in robot research.The natural supple movement characteristics of the human body provide a perfect reference for the study of the humanoid robot's humanoid movement.The human joint motion has nonlinear time-varying characteristics.How to qualitatively and quantitatively analyze and estimate the motion information of human joints is the premise and basis for realizing the humanoid motion control of robots.In this paper,the muscle-driven upper limb elbow movement state is studied,and the filtering algorithm is used to estimate the elbow joint movement status information,which lays the foundation for the robot to track the human joint movement status.Based on the analysis of the human body's nerve signal on muscle force or moment,this paper establishes a neuromuscular skeleton model of human elbow joint movement under the action of biceps and triceps.Because of the suppleness of the continuous motion of the elbow joint driven by the muscle,the suppleness of the joint depends on the suppleness of the muscle.Therefore,the flexible characteristics of the muscles at the elbow joints of the human body are analyzed and studied to realize the quantitative calculation of the joint stiffness values of the elbow joints.The sEMG and joint angle acquisition platform was built for the muscle groups that caused the elbow joint movement.By designing a signal acquisition experiment scheme,information acquisition of surface EMG and elbow angle is realized,which provides a reference for the validation of subsequent estimation algorithms.The state space model of the elbow motion mechanism model was established.The experimental data measured by the experiment were reference values.The unknown parameters in the state space model were identified by LM(Levenberg-Marquard)algorithm.For the established state space model,Extended Kalman Filter(EKF)algorithm and Unscented Kalman Filter(UKF)algorithm are used to estimate the angular displacement and angular velocity of human elbow joint motion.The results show that two algorithms areused to estimate the motion state of the elbow.The results of the two filtering algorithms are compared.The results show that the estimated effect of UKF is better than that of EKF.The stiffness parameter and motion state of the elbow joint are jointly estimated.By comparing with the UKF algorithm,it can be seen that the joint estimation method is further improved in the estimation accuracy.