The invention relates to the field of 
artificial intelligence, in particular to a ventilation man-
machine asynchronous detection model training method and device based on DQN 
reinforcement learning, and the method comprises the steps: obtaining a capacity 
data segment during 
breathing and ventilation; preprocessing the capacity 
data segment to obtain training data and 
test data which can be used for DQN 
reinforcement learning; based on the training data and the 
test data, constructing a scene problem suitable for applying DQN 
reinforcement learning processing; describing or depicting the scene problem; and constructing a specific DQN network, setting training parameters of a reinforcement learning model, and training a DQN reinforcement learning model. A man-
machine asynchronous detection problem of 
breathing ventilation is converted into a scene problem, so that the man-
machine asynchronous problem can be identified or detected by using a reinforcement learning scheme. A 
processing method of 
breathing ventilation man-machine asynchronous detection is expanded, the problem of automatic detection of man-machine asynchronous events occurring in the breathing ventilation process is solved, the monitoring burden is effectively relieved, and the 
nursing efficiency of work is improved.