Research And Development Of Module-assembled Emg-assiited Rehabilitation Intelligent Wheelchair

With the deepening of the aging of the social population and the increasing number of disabled and weak people,more and more institutions at home and abroad have focused their research on the intelligent wheelchairs necessary for such people to live and travel,especially the modules of intelligent wheelchairs.With the rapid development of artificial intelligence,precise control strategies,active safety protection,and identification of patient rehabilitation movement intentions are not only the research hotspots of intelligent wheelchairs,but also active health and Application hotspots in the field of health care.In this paper,a set of underlying control system is designed with STM32 as the core.Six functions including manual joystick control,control mode switching,bottom active safety,QT host computer control,remote wireless communication,and power management are designed,making the bottom motion control system of the smart wheelchair more secure and compatible.In addition,in order to meet the control needs of different disabled people,in addition to manual joystick control,this paper also designs four control interfaces such as automatic control,voice control,and EMG control.In the fully automatic control interface,the ROS platform is first built on the wheelchair,and the feasibility of the wheelchair model and the effectiveness of the A*and DWA（dynamic window method）algorithms are verified through simulation,and then the intelligent wheelchair is realized in the real environment.Fully autonomous navigation and local obstacle avoidance.In the voice control interface,this paper designs an online voice recognition system based on the i FLYTEK voice database and an offline voice recognition system based on the YS-LDV7 module.Both methods can ensure the recognition accuracy of voice commands is above 75%.In the EMG control interface,firstly,a distributed wearable EMG instrument was prepared to collect and transmit surface electromyography signals（s EMG）,and secondly,the signal acquisition points were determined by studying the walking gait of people and the leg health recovery strategy was set.Then design experiments to collect the original EMG signal and cut it into data samples of a single action through data processing;on this basis,a CNN EMG recognition model is built,and various indicators of the model are evaluated,and finally real-time control is designed.The system access model realizes the transformation from s EMG signal to control command.To sum up,this paper refits the electric wheelchair,designs a smart wheelchair prototype composed of multiple modules and has health recovery function,and conducts several sets of experiments on the wheelchair,including the global driving and Local obstacle avoidance experiment,voice control experiment,s EMG signal acquisition experiment,etc.The experimental results show that the intelligent wheelchair designed in this paper has strong feasibility.On the basis of ensuring the safe and convenient travel of different groups of people,it can also meet the health recovery needs of some users,and can integrate travel and rehabilitation,which is of great importance.Application value and social benefits.