Recognition Of Motion Intention Based On Biosensor For Rehabilitation Robot

Stroke is recognized as the second leading cause of death globally and the main cause of long-term disability in adults.Worldwide,approximately 15 million people suffer from stroke each year.Foot drop is part of the common diseases after stroke.In order to treat the hemiplegia caused by foot drop,the Fugl-Meyer motor score scale and Barthel index are used clinically to analyze the patient’s ability to perform and evaluate the patient’s onset period.Immediate rehabilitation training in the early stage of the disease can maximize the gait close to an average person.Treatment methods include massage training,electrical stimulation training etc.In recent years,the introduction of exoskeleton robots has added new solutions to rehabilitation treatment,but the current cost of exoskeletons is expensive,and the intelligent interaction between the exoskeleton robot and the wearer Sex has yet to be improved.In the aspect of intelligence research,the sensors selected by scholars include human muscle electrical signals,EEG,and attitude signals,and the selection of sensing signals is relatively simple.To this end,this research is aimed at the motion intention recognition and robot application of multi-source sensing signals,and by introducing biological signals and physical signals to predict the joint angle of the human body,the exoskeleton assists the rehabilitation training of hemiplegic patients.Firstly,according to the unilateral hemiplegia walking style of stroke patients,an efficient and economical modular exoskeleton is designed.In terms of mechanical structure,a lead screw and crank structure are used to enhance the coupling in the mechanical components.The electronic control adopts STM32F4 series chips to ensure the real-time performance of the task system.Choose disc motors and brushless DC gear motors to provide extra power for human movement.Secondly,in order to solve the problem of how to generate rhythmic and symmetrical movements during the rehabilitation of the lower limbs of hemiplegic patients,a model of motion intention recognition based on vibroarthographic signals and prediction of contralateral joint angles based on multi-source biological signals is proposed.Gesture recognition is achieved using vibroarthographic signals.Then,sound signals are introduced into the lower limbs,vibroarthographic sensors are connected to the affected legs,and Surface Electromyography sensors are connected to the healthy legs.The corresponding signals are used to estimate hip,knee,and ankle joint angles of the affected leg,and a temporal convolutional network-based algorithm was introduced to predict the contralateral lower extremity joint angle during human motion.Finally,the purpose of this paper is to improve the comfort of patients wearing exoskeletons for rehabilitation training.The coupling of the exoskeleton and the human body model is regarded as a system,the kinematics and dynamics formulas of the human body are established,and the controller model based on admittance control and proportional-differential control is designed.By measuring the interaction force between exoskeleton components and limbs,the virtual admittance model is used to correct the inherent characteristics of the system,thereby improving the flexibility of the exoskeleton.