Design And Application Of Smart Sensor System On Lower-limb Exoskeleton

The lower-limb exoskeleton,a medical equipment used to help patients with paraplegia and the elderly regain the ability to walk,could be worn for walking training for the purpose of rehabilitation and activity recovery.Due to the disability of the patients,the secondary damage should be avoided,so the safety and comfort of lower-limb exoskeletons are crucial.Based on the research of function requirements and actual usage of lower-limb exoskeleton,this thesis proposes a set of intelligent sensing system for lower-limb exoskeleton robots to ensure the safety of lower-limb exoskeletons and enhance the comfort.By analyzing the functional requirements of patients and combined with the actual research progress of the lowe-limb exoskeleton platform,the scheme of the sensor system is confirmed and the selection of each sensor is completed.The corresponding circuit could then be designed and produced,followed by building the hardware platform of the circuit to accomplish the sensor data collection.According to the principle of sensor measurement and characteristic curve of signal input and output,sensors are calibrated.The accuracy of sensors has been significantly improved after precise calibration and fully meets the current requirement of the lower-limb exoskeleton system.IMU measurement platform is designed and produced and then used to measure 3D posture data of lower-limb exoskeletons during walking and the length and height of steps based on the principle of inertial navigation.The platform gives real-time feedback to the main control system of lower-limb exoskeletons to collect gait data for real-time adjustment.The lower-limb exoskeleton gait is divided based on timing and magnitude of force on each sensor.Based on the accurate data of pressure sensors the coordinates of center of plantar pressure Co P is calculated using ZMP,and provides strong theoretical and practical support for the stability of the lower-limb exoskeleton during walking.To recognize of the moving intention of the wearer,based on the sensor system,data of each sensor are collected when lower-limb exoskeletons are worn by wearers in the walking test,which is used as training sample that is learned using support vector machine.Furthermore,the wearer's moving intention is identified using the learning results and is tested by experiment.The experimental results show that the identification of the moving intention of the wearer performs well by this method,which is able to reduce the delay from receiving command to completing it when the lower-limb exoskeleton is controlled manually and is able to improve the comfort.