Research On Action Recognition And Behavioral Predictive Method Of Labor-type Exoskeleton Robot

Based on the current high labor intensity and low work efficiency in the agricultural field,workers are provided with an exoskeleton robot dedicated to agricultural labor,which is applied to field management and harvesting in the agricultural field.Labor intensity,extension of working hours,reduction of labor force,increase of labor productivity,increase of farmers' income,reduction of diseases and physical injuries caused by labor,promotion of healthy development of agriculture,and harmony and stability in rural areas are important for the protection of China's economic security.This article mainly carried out a preliminary study on the action recognition and behavioral prejudgment of labor-type lower extremity exoskeleton.The specific content is as follows:When the lower extremity exoskeleton mechanism assists the human body in exercising,it needs to cooperate with the human body to complete various movements.Therefore,it is necessary to ensure that the joint structure and the movement characteristics of the exoskeleton lower extremities are consistent with the human body.First,a labor-based human lower limb exoskeleton mechanism was established,which can meet the coordination and followability,and the stability of the exoskeleton mechanism can be ensured through the control of instability.The kinematics modeling and simulation analysis of the established labor-type exoskeleton exoskeleton were carried out.Through the analysis of kinematics forward and inverse solutions,the analysis of Jacobian matrix and singularity,and the judgment of instability,a stability control strategy is proposed.Through the analysis of its working space,the stability of this model is verified by Matlab simulation.,and then through ADAMS simulation to further prove that this organization in the human body when the load is good.When performing motion recognition,first wear various acceleration sensors on the wearer to fuse the collected data and extract feature signals,then use RBF neural network to classify and recognize them,and thereby determine the labor actions of the human body.After preliminarily recognizing the behavior of the exoskeleton of the human lower extremity,the Kalman filtering algorithm is used to predict the behavior of the lower extremity exoskeleton robot.The exoskeleton model is simplified as a two-bar linkage mechanism for kinematics analysis.Then,the PD control strategy is used to ensure the stability of the system.Through simulation analysis,the followability of the exoskeleton mechanism is verified and the human behavior is accurately predicted.