Study On Voluntary Functional Electrical Stimulation Control System For Elbow Joint Motion

Due to brain or nervous system injury,it is difficult for patients with limb dysfunction in completing the desired action.Stroke and spinal cord injury,are the major causes that render limb dysfunction.The incidence of such diseases,nowadays,with the aging of the population and the development of urban transport,is keeping rising,resulting in an increasing trend of dysfunction.Moreover,limb dysfunction could impose seriously effects on their daily lives and burdens on their families.Therefore,it is imperative to develop an effective limb 'rehabilitation technology towards such patients.Functional electrical stimulation is an effective rehabilitation training method for the limb dysfunction which widely used in the recovery and reconstruction of limb motor function.At home and abroad,researches about FES system for the reconstruction of limb movement function have evolved from the open-loop training system to the closed-loop training system and from the passive training system to the active training system.Lightened by the development trend of FES system,we designed a FES system with initiative and closed-loop control of elbow by taking the elbow as the research object,integrating the patient's voluntary intention control,and combining the actual angle feedback.Specific research contents are as follows:(1)In order to extract voluntary intention,the surface EMG signal components under voluntary control and electrical stimulation were experimentally analyzed.An online filtering algorithm combining blanking and template substracting was proposed to eliminate the components of electrical stimulation artifacts in surface EMG signals.Then,HP filter algorithm was used to filter the M wave components in the surface EMG signals.Finally,the desired voluntary EMG signal were obtained.(2)In order to establish the model of elbow motion under electrical stimulation,a dynamic neural network model based on Hammerstein model structure with feedforward and feedback links was proposed.The experiment of elbow joint movement induced by electric stimulation was designed,and the experimental data were collected as network training data.The genetic algorithm was used to train the neural network,and the optimal model parameters were obtained.Finally,the model of electrical stimulation and elbow joint motion angle was obtained.The model was verified with test data,which showed the validity of the model.(3)In order to achieve accurate control of elbow movement angle,the FES system using iterative learning control was investigated.The FES closed loop simulation system was completed in the Simulink toolbox of MATLAB,and the effectiveness of the simulation system was analyzed.Simulation of ILC-based FES system was implemented in the Simulink toolbox of MATLAB using proposed neural network model as control plant.Proposed FES system was also verified experimentally.The simulation and experiment results both illustrate the effectiveness of ILC.The system was applied to the experimental verification of the actual elbow stimulation to illustrate the effectiveness of the iterative learning system.The voluntary will signal extraction algorithm presented in this thesis is fast and effective,which is capable of extracting the movement intension for voluntary controlled FES.The established neural network model of elbow joint motion under electrical stimulation can well reflect the motion characteristics of elbow joint under electrical stimulation.The iterative learning control of the elbow joint FES system exhibits a excellent control effect.