Design And Implementation Of Obstacle Avoidance Arm Movement Rehabilitation System Based On Motor Imagery

Brain computer interface technology,as a tool to assist,enhances and restores the motor ability of the clinical people,and has been studied as an auxiliary treatment for the rehabilitation of stroke hemiplegia patients.BCI technology can help to improve and restore the motor sensory ability of clinical patients.For example,through BCI system to control the robot arm,to fully induce and promote brain plasticity,to mobilize the willingness of hemiplegic patients to actively exercise,so that patients can actively carry out rehabilitation training,so as to obtain good results and restore the motor ability.Because of the movement obstacle of patients' arm,it is easy to collide and cause injury when the robot arm is training.This thesis aims to restore the ability of arm movement and design an on-line robot rehabilitation system based on motion imagination,which can preprocess,extract features and classify patterns of the collected EEG signals,and then control the robot arm and the infrared sensor to detect the distance between the arm and the surrounding objects to prevent arm collision respectively.At the same time,the face recognition function is added to further strengthen the practical operation function of the rehabilitation system.This thesis mainly divided into two parts: the processing of EEG signal and the design of mechanical arm system(1)Collection and analysis of EEG signalsWe mainly introduce the EEG signal acquisition process based on Open BCI platform,collect the corresponding EEG signals of motion imagination,and through preprocessing,which is based on the Improved EMD combined with WPT and CSP methods,and support vector machine to classify the patterns of the signals.The average classification accuracy rate of all 9 subjects is95.9%.The results show that the feature extraction method is feasible and effective.(2)Design of on-line rehabilitation robot systemThis part introduces the control module of the robot arm,the software module of the computer and the hardware instruction receiving and executing module of the lower computer with six degrees of freedom.The PC transmits the command after the EEG signal classification to the robot to control the bending and extension of the manipulator online.When the arm is moving,the signal from the infrared sensor on the arm is used to avoid obstacles.Finally,combined with the human face recognition module,the function of a mechanical arm is useful in real life.When the cameracatches the human face,the mechanical arm extends the water cup to the face direction to complete the auxiliary water drinking action.The thesis analyzes the results of the on-line control experiment of brain computer interface system to the robot arm,which shows that the research in the paper can process the EEG signal well and control the robot arm through the classified instructions to achieve the expected action,help patients to carry out rehabilitation training.This paper provides a new idea for the practical application of brain computer interface system in the rehabilitation field.