Research On Multi-degree-of-freedom Manipulator Control Method Based On BCI Technology

With the development of science and technology,the study of brain science and human brain intelligence technology is getting deeper and deeper.As the most representative emerging technology in brain science research,brain-computer interface technology based on movement imagination has attracted more and more attention from domestic and foreign researchers.This technology can bypass the peripheral nerve and muscle tissue of human brain,collect the EEG signals corresponding to each movement intention in human brain,analyze and process the EEG signals of movement imagination and convert them into control instructions,respectively control the external equipment to complete the corresponding tasks,so as to achieve the purpose of communicating with the external environment.Brain-computer interface technology is promising for patients with limb movement disorders or neuromuscular tissue damage but sound thinking.Most of the previous international brain-computer interface competitions were based on the motor imagery of left and right hands,feet and tongue,while there were few studies on the movements of different joints of human limbs.Based on the existing four kinds of motor imagery,this paper designed three kinds of motor imagery experimental paradigms based on upper limb shoulder,elbow and wrist.In order to achieve the control of external devices through brain-computer interface system,the in-depth analysis of EEG signals is essential.In this paper,for the three kinds of experimental models of motor imagery designed by the subject,the EEG signal processing method based on Wavelet Transform Analysis,EMD-CSP and GA-TWSVM is proposed,and will eventually be applied to the control of external mechanical arm.Specific research contents are as follows:(1)Research on experimental paradigm of motor imagery.Based on the current research on the brain-computer interface technology of motor imagery,most of which focus on the motor imagery of left and right hands,feet and tongue,this paper designs three kinds of motor imagery experimental paradigm-forward flexion of right arm,forward flexion of elbow and outward rotation of wrist.Eight subjects were selected to collect three types of motor imagery EEG signals as the experimental data set in this paper.(2)Research on methods of EEG signal preprocessing and feature extraction.1)Aiming at the experimental paradigm of motor imagery designed in this paper,the common average reference was used to filter the original EEG data,and then the frequency bands related to motor imagery were extracted.2)Power spectrum analysis is used to verify the effectiveness of the preprocessing,and feature extraction method of three kinds of motor imagery based on EMD-CSP is proposed,to analyze the intrinsic mode function obtained by empirical mode decomposition of the preprocessed signal,and input the time-domain information corresponding to the frequency band of motor imagery into the common spatial pattern,which further improves the feature extraction ability of the common spatial pattern.(3)Research on recognition of motor imagery pattern.Twin support vector machines optimized based on genetic algorithm are proposed to identify and classify three kinds of motor imagery.Genetic algorithm is used to optimize the parameters of the twin support vector machine,and the experimental data are processed by genetic algorithm to optimize the twin support vector machine,least squares support vector machine,BP neural network and extreme learning machine.By comparing the recognition rate of each method,it is proved that the algorithm proposed in this paper is superior to other existing methods.(4)Design of manipulator control system.The overall framework of the system is designed,the function of each module of the system is introduced in detail,the hardware and software of the system are designed,and the details of the system are explained.Finally,check the performance of the control system,design the experiment to test,analyze the test results,draw conclusions.It is proved that the proposed algorithm has certain practical value for the control of manipulator,which lays a foundation for the realization of more types of movement control of manipulator.