Research And Design Of Smart Home System Based On Brain-computer Interface

With the rapid development of today's technology,people's demand for intelligence in smart home is increasing.In recent years,people have achieved some practical results in brain computer interface(BCI)technology and electroencephalogram(EEG)research.How to realize the"brainwave"control of home equipment is a new research direction in today's smart home.Only when the BCI technology is applied to the smart home system and the human"mind control"is accurately reflected in the smart home system,the smart home system can be more intelligent.In this thesis,based on BCI technology,the EEG signals are collected,preprocessed,extracted,classified and recognized,and finally transformed into commands to control smart home devices,thus the brain control function of smart home system was finally achieved.The main research contents of this thesis are as follows.(1)Acquisition and preprocessing of EEG signals.Through event-related desynchronization/event-related synchronization(Event-related Desynchronization,Event-related Synchronization,ERD/ERS)analysis,the C3,C4,F3,and F4 electrode EEG signals corresponding to the frontal brain area and central brain area are determined as the on and off signals of household equipment(table lamps,fans)for research.This thesis uses the laboratory EEG signal acquisition device Emotiv EPOC+EEG to design an acquisition scheme for EEG signal acquisition.Aiming at the problem of noise interference and distortion of the collected original EEG signals,this thesis adopts an adaptive sample entropy wavelet threshold function denoising method,and constructs an adaptive threshold function based on sample entropy.Compared with traditional methods,it has a better performance.Simulation results show that the method is effective.(2)Feature extraction,classification and recognition of EEG signals.In terms of feature extraction,this thesis introduces three methods:CSP,AR model,and PSD for feature extraction and simulation analysis.In terms of classification and recognition,this thesis introduces Bagging algorithm and Adaboost algorithm for classification and recognition.In view of the shortcomings of the above means,an improved Adaboost classifier algorithm is proposed in this thesis,it clusters the training data set and updates the coefficient ?_m of the low-precision classification model G_m(x)to ?_m.After simulation,it is found that the accuracy of this algorithm is better,and the feature values extracted by the CSP algorithm are processed by the improved Adaboost classifier algorithm.The signal classification accuracy rate is the highest,so the CSP algorithm is used to extract the features of the EEG signal in this thesis.(3)The design of intelligent home system.This system includes software design and hardware design,the software design includes four parts:gateway platform programming,terminal node programming,host computer programming,and Web cloud platform programming,and the hardware design includes gateway platform(main control module,Wi Fi module),ZigBee module,terminal node module,and peripheral circuit.The STM32F103C8T6 was selected by main controller of the gateway platform to complete data processing and control of terminal equipment.Data is interacted between the home system and the Web cloud platform through the Wi Fi module.The wireless networking of each node is realized through Zig Bee module.The circuit design of terminal node module includes sensor node and control node.(4)System's experiment and analysis.The smart home system based on BCI is experimentally tested,the effectiveness of the adaptive sample entropy wavelet threshold denoising algorithm,the improved Adaboost-CSP classification algorithm and the feasibility of the designed system were verified by experimental results.