Research On Nursing Bed Brain Control Methods Based On Brain-computer Interface

In this paper,the brain control method for nursing bed based on brain-computer interface is taken as the research object,and the theoretical methods such as preprocessing,feature extraction and recognition are studied.The steady-state visual evoked potential in brain-computer interface are used as control signals,a nursing bed brain control system based on steady-state visual evoked potential is constructed.The main innovations as follows:(1)The steady-state visual evoked potential stimulation paradigm of the nursing bed brain control method is established.Based on the 10-20 international electrode placement standard,the accurate placement point of the steady-state visual evoked EEG signal acquisition electrode is determined.Aiming at the defects of the traditional visual stimulator,which has low visual evoked rate and easy to cause visual fatigue,a contrast experiment of different color flash stimulation and graphic stimulation on a black background is performed.By analyzing the experimental results,it's determine that the checkerboard visual stimulation mode is used as the stimulation mode of the steady-state visual evoked potential stimulator.(2)Aiming at the wavelet coefficients are not continuous after the hard threshold function denoising and the constant deviation between the wavelet coefficients and the real wavelet coefficients after the soft threshold function denoising.An improved threshold function is Proposed,the wavelet coefficients after denoising by the improved threshold function gradually approaches the real wavelet coefficients,and the improved threshold function is used to the EEG signal denoising,thereby improving the signal-to-noise ratio of the EEG signal.(3)Aiming at the problem that the early convergence speed is fast,the later is slow,and it is easy to fall into the local optimum during the optimization process of artificial fish swarm algorithm,an improved artificial fish swarm algorithm(IAFSA)is proposed.The BP neural network is optimized by this algorithm,and the BP network model based on improved artificial fish swarm algorithm is obtained.The problem of BP neural network randomly selects initial value is solved and global optimization is realized.The verification of the above theoretical methods is achieved by constructing a complete nursing bed brain control system.Experiments show that the BP network optimized by the improved artificial fish swarm algorithm is used to identify the EEG signals.Compared with the basic BP network and the BP network optimized by the traditional artificial fish swarm algorithms,the recognition rate increased by 4.7%and 2.2%respectively;The correct recognition rate of the nursing bed control instruction reaches more than 88.89%,and the induced operation time of each control instruction is less than 8s,which proves the correctness of the algorithm studied in this paper and the reliability of the constructed system.