Research On Human-computer Interaction Control Based On Eeg Signals

The emergency response process of the traditional driving mode with hands and feet coordination takes a long time and is prone to fatigue,which can cause unnecessary dangers.The assisted driving method based on electroencephalogram signals can detect the fatigue state of the human brain,reduce the reaction time,as a supplement to manual driving,and avoid unnecessary accidents.With the continuous development of artificial intelligence and brain-computer interface technology,neural network computing that mimics the operation of human neurons is getting closer to the human brain computing model;brain-computer interface technology provides a bridge for electroencephalogram signals to directly communicate with the outside world.The combination provides a feasible solution for the human-computer interaction control of assisted driving based on electroencephalogram signals,and maximizes the safety of driving.This paper studies the human-computer interaction control method based on EEG signals,and the main work are as follows.In order to solve the problems of the traditional electrode,such as low sensitity,poor antiinterference ability,and inadequate fit with the skin,resulting in weak signals and complex interference signals in the collected data,this paper designed and developed a micro-pillar electrode assembly based on flexible materials.First,the flexible micro-pillar substrate is prepared by the electric field-assisted imprinting method,then the silver electrode is prepared by ion implantation and magnetron sputtering,and finally the silver chloride is prepared by 3D printing chlorination,which is uniformly sprayed on the surface of the silver electrode.Experiments shown that the flexible micro-pillar electrode retains the advantages of traditional electrodes and can better fit the skin,improve the signal quality,reduce the interference of external signals,and achieve high-precision electroencephalogram signal acquisition.Aiming at the low accuracy of electroencephalogram signal classification and recognition of traditional convolutional neural networks and long-short-term memory networks,this paper studies and implements a multi-scale fusion convolutional neural network based on the attention mechanism and bidirectional long-short-term memory network parallel electroencephalogram Signal classification method.The multi-scale fusion convolutional neural network is used to extract the information within and between brain regions,avoiding the loss of information when extracting features in advance.The two-way long and shortterm memory network considers the time steps before and after the current state,and solves the long-term dependence problem in the network model.The attention mechanism module in the network is used to improve the expressive ability of the framework,increase the sensitivity of the model to information characteristics,and improve the accuracy of the network.The results of the multi-scale fusion network are combined with the results of the long and short-term memory network to complete the electroencephalogram signal classification task.Tests on public data sets prove that this method is superior to other comparison methods in electroencephalogram signal classification.In order to verify the effectiveness and practicability of the electroencephalogram acquisition module with flexible micro-pillar electrodes as the core and the electroencephalogram classification algorithm based on deep learning,this paper studied and implemented an unmanned platform control software based on electroencephalogram signals.The electrical signal is preprocessed by steps such as filtering,down-sampling,re-reference and independent component analysis,as the input of the electroencephalogram signal classification network.After the classification,the control instructions corresponding to the electroencephalogram signal are designed.This paper combines Car Sim and Simulink to simulate the actual driving situation of the vehicle,and realizes the horizontal and vertical joint control of the vehicle based on the bioelectric signal brain-computer interface in the virtual environment.At the same time,the designed control scheme was tested on the unmanned car,and the corresponding interactive software was designed,which realized the control of the unmanned car based on electroencephalogram signals,and verified the feasibility of the principle.