Motor imagery electroencephalogram signal classification method and device and storage medium

The invention discloses a motor imagery electroencephalogram signal classification method and device and a storage medium, and the method comprises the steps: carrying out the band-pass filtering processing of source subject motor imagery electroencephalogram signal data according to a preset Hertz range, and obtaining a source domain; training a preset deep learning model to obtain a source domain model; initializing an initial target domain model; performing band-pass filtering processing on the motor imagery electroencephalogram signal data of the small batch of target subjects to obtain a target domain; replacing an original batch normalization layer of a feature extractor in the initial target domain model with a dynamic mixed batch normalization layer to obtain a first target domain model; calculating features of the target electroencephalogram signal data after the target electroencephalogram signal data are input into the dynamic mixed batch normalization layer; replacing a linear classifier in the fi