Robot Wheelchair Control System Based On The Interactive Fusion Of Motor Imagery EEG And Head Posture

With the increased attention paid to the elderly and disabled groups,the need to meet the travel needs of these groups has become a pressing social issue.Robotic wheelchairs are the application of technologies related to mobile robots to ordinary electric wheelchairs and have broad application prospects.As one of the key technologies in robotic wheelchairs,the human-machine interface has a crucial impact on the harmonious operation between the user and the wheelchair.The head posture interaction control can help the user with the human-machine interaction of the wheelchair through intuitive head movements,but is prone to secondary injuries to the user due to muscle fatigue.Brain-machine interfaces are an emerging technology,one of whose ultimate research goals is to enable control of external devices by decoding the user’s brain activity in real time,which can help users with only the ability to think to reconnect with the outside world.On this basis,this paper designs a robotic wheelchair control system based on the interactive fusion of motor imagery EEG and head posture,the main elements of which are as follows:(1)This paper implements continuous control of a robotic wheelchair based on motor imagery EEG.In contrast to traditional EEG control methods,which mainly provide discrete control commands for the wheelchair to move forward,backward or turn in place,this method decodes the subject’s sensorimotor rhythms,extracts the signal features most relevant to imagining different movements,and combines them into a control signal to control the motion of a virtual cursor.The motion of the cursor is then mapped in real time to the wheelchair control interface and output as speed control commands for the left and right wheels of the wheelchair,ultimately achieving continuous control of the wheelchair.(2)In order to train subjects to control the cursor using sensory-motor rhythms,a series of step-by-step cursor control exercises were designed,starting with one-dimensional movements of the virtual cursor and expanding to two-dimensional movements,then offsetting the position of the target block and reducing the size of the target block.The series of step-by-step training sessions were designed to help the subjects improve their ability to regulate their sensory-motor rhythms,and eventually to precisely control the virtual cursor movements to manipulate the robotic wheelchair.(3)The fatigue state of the subject’s neck muscles was used as the switching condition to establish a robotic wheelchair control system integrating motor imagery EEG control and head posture interaction control.The fatigue state of the subject during wheelchair control is assessed by collecting the surface EMG signals from the subject’s neck muscles,calculating the integrated EMG value and the average power frequency as two evaluation indicators for combined amplitude-frequency analysis,and switching between the two control modes according to the assessment results.