EEG control method for fire extinguishing UAV based on improved subband filter bank

This paper designs a cloud-brain control method and system for fire-fighting based on sub-band pass filter group for the problem that firefighting troops can't quickly grasp the fire point information when they rescue fires in high-rise buildings. The key technology for controlling UAV using EEG signals is the acquisition, feature extraction and classification of EEG signals. First, we use the Mexican straw hat wavelet transform to decompose and reconstruct the original brain waves. Secondly, based on the pre-processed brainwave signal, a redesigned sub-bandpass filter bank filtering method is proposed, which greatly improves the speed of feature classification. Thirdly, combined with the filter bank, the single frequency range filtering method in the common spatial paternal mode is improved, and the optimal passband is found according to different brain wave characteristic signals for feature classification. Experiments show that the aircraft control application is developed based on Android system, integrating EEG signal processing, information coordination and flight control command functions to achieve convenient operation. Experiments show that the control method designed in this paper improves the accuracy of feature classification by 16% compared with Fast Fourier algorithm and Discrete Fourier aldronegorithm when processing visually induced EEG signals. The accuracy of the average algorithm and the superposition average algorithm are improved respectively. 9% and 4%, the control strategy is effective and stable.