Research On Ankle Rehabilitation Robot Control And Intellisense Based On Multi-information Fusion

The development of an aging society has brought about the need for rehabilitation treatment of stroke patients.The ankle joint is the distal joint of the human body,and the ankle rehabilitation plays an important role in the walking function Ankle rehabilitation plays an important role in the walking function testoration of patients with lower limb dyskinesia.Existing ankle rehabilitation training relies on the manual operation of the therapist with passive training,which affects the rehabilitation efficiency of patients and cannot provide accurate training programs.Therefore,this paper proposes a new ankle rehabilitation robot that can imitate the therapist’s manipulative techniques.This robot can achieve self-rehabilitation training for patient that combines multiple training modes by uses the patient’s healthy side to drive the affected side.And this paper put forward the ankle joint motion perception system for patients to provide a basis for the specified training program.Firstly,the research status of ankle joint rehabilitation robots at home and abroad were analyzed in this paper and the related research results of human motion perception system were introduced.By analyzing the characteristics of the existing robot system,this paper proposes to design a rehabilitation robot with mirror mechanism which can simulate the technique of the therapist.At the same time,the patients’ motion perception system is proposed by integrating a number of different types of sensors,and it can objectively evaluate patients’ rehabilitation status and make reasonable training plans.Based on the analysis of the clinical rehabilitation treatment technique and the sports therapy,the author innovatively proposes the movement mode of the combination training of the healthy side and affected side on the basis of the traditional active and passive training method,and the design requirements of the robot system is proposed.Then the whole system of the robot is introduced.From the perspective of bionics,the physiological structure and motion characteristics of the patient’s ankle joint is analyzed,and the mechanism design scheme of the robot by simplifying the motion model of the ankle joint is proposed.The robot adopts symmetrical structure design,which connects the base,the internal/external rotation mechanism,the inversion/eversion mechanism,and the dorsiflexion /plantarflexion mechanism,and the rationality of the mechanical design is verified through kinematic analysis.The hardware system of the robot is built by selecting the appropriate motor,driver,industrial control computer and sensor.The peripheral control circuit is design and the control cabinet is built.The robot control system is introduced based on the robot system design requirements,and the GUI interface is built based on C++ language.The automatic training system and the manual training system are proposed for doctors and developers.Then,based on the clinical rehabilitation needs of patients,three training modes consist of the passive training,the active training and the combination training of healthy side and affected sides are proposed.The position control strategy,torque control strategy and follow control strategy are introduced also.The PID control module of Simulink simulation model is built In MATLAB for speed response experiment.Then the feasibility of the proposed control strategy is verified by single-joint single-period motion experiment.Then the design requirements of multi-information fusion perception system are proposed by extracting the key parameters of ankle joint motion perception.The force of the patient’s internal/external joint and the dorsiflexion /plantarflexion joint are collected by the pressure sensors,and the force of the patient’s internal and external rotation joint is collected by the torque sensor.The wearable inertial sensor was used to measure the range of movement of the patient’s joints.The EMG sensors are used to collect the changes of EMG signals of the tibial anterior muscle during training.Finally,based on the built hardware system,a motion state perception strategy was proposed.Finally,the motion state perception strategy is proposed based on the hardware system.Finally,the robot experimental platform is built and qualified clinical experimental subjects are recruited.The healthy people are invited to carry out the active and passive training and combined training on the ankle joint rehabilitation robot platform,and the control performance of the robot is verified by analyzing the errors of speed and position.The correctness of the perceptual system is verified by collect the data of the joint force,the range of motion and EMG signal.The patients’ movement state is analyzed based on the relevant data,and the patients’ recovery evaluation model is further put forward.