Gait Planning And Control Of Exoskeleton Robot Based On Intention

Exoskeleton robot is a kind of intelligent device,which mainly helps people with lower extremity paralysis to walk by simulating normal people's movement.The exoskeleton robot integrates sensing,control,information,movement,and computation,and its function is realized utilizing multi-part cooperative "intelligent" control.An important direction in the research of exoskeleton robot is to switch different gait according to different motion intention,to realize the planning of active control gait by recognizing the motion intention of the wearer,and to help people with lower-limb paralysis to realize walking.Aiming at the part of the wearer's motion intention recognition,this research proposes to use BP neural network to decode muscle electrical signals to complete the recognition of human motion intention,plan the motion gait through the result of intention recognition,and analyze the stability of the planned gait.To solve the problem of moving gait planning,we first need to divide the gait phase.In this research,we construct an experimental platform based on SIAT rehabilitation exoskeleton robot to collect surface electromyography(s EMG)signals of 6 channels under normal walking by exoskeleton wearers and perform feature extraction.Vicon-Nexus equipment is used for data acquisition and verification.Experimenters were surface electromyography equipment to walk in the three-dimensional force measuring platform to detect the center of gravity,joint trajectory,motion trajectory,and motion pressure during the travel.Calculated the motion trajectory of the center of gravity of the experimenter,compared with the stability and minimum pressure of the center of gravity in the Vicon motion capture system,and obtained the best gait motion trajectory.Using a support vector machine(SVM),LSTM,BP neural network to divide the s EMG signal feature vector to realize the phase recognition of the wearer's four kinds of motion gait.Secondly,the sagittal motion trajectory of the joint is approximated as a sine function combined with different gait phases to construct a motion model.Finally,the planning of moving gait is completed according to the planned motion model.To solve the problem of cooperative control of the exoskeleton robot,this research constructs a motion gait system based on human motion intention to control the exoskeleton robot in real-time.By establishing an IP address and port communication,the s EMG signal acquisition module,intention recognition module,and exoskeleton control gait switching module are connected to realize the planning of exoskeleton equipment to control motion gait switching in real-time by identifying s EMG signals.Experimental results show that the s EMG signal decoding human motion intention control exoskeleton robot gait switching has good accuracy and real-time,helps patients to complete rehabilitation training more safely and quickly,greatly improves the performance of exoskeleton robot human-computer interaction,promote the development of intelligent equipment.