The invention belongs to the technical field of
robot control, and discloses a
central nervous system-based
exoskeleton control method and
system, and the method comprises the steps: collecting a central nervous
signal in a
human body, and carrying out the preprocessing of the central nervous
signal, and obtaining a multi-dimensional nerve
feature set; performing decoding analysis on the multi-dimensional neural
feature set by utilizing a preset dynamic neural intention decoding model to obtain a
human body motion intention, and generating joint motion parameters and a driving force control instruction of the
exoskeleton robot by utilizing a collaborative model; and the
exoskeleton robot is controlled to conduct joint movement, closed-loop feedback is constructed, and the driving force control instruction is adjusted in real time. According to the method,
information loss or
distortion caused by damage of a central nervous
conduction pathway is effectively avoided, the adaptability to a central nervous injury patient is improved, the decoding accuracy is improved, the
motion synchronization error is reduced, the control fluency and the man-
machine coordination of the
exoskeleton robot are improved, and the control safety of the
exoskeleton robot is ensured.