Depth state space model-based interpretable process monitoring method and related equipment

The invention is suitable for the technical field of industrial process monitoring, and provides an interpretable process monitoring method based on a depth state space model and related equipment, and the method comprises the steps: carrying out the modal division of industrial process data samples at a plurality of historical moments, and obtaining a modal label corresponding to each industrial process data sample; segmenting the industrial process data sample with the modal label to generate a training sample set; constructing a depth state space model; hidden variables of the depth state space model comprise an understandable discrete part and an ununderstandable continuous part; taking logarithm likelihood of maximum observation and classification cross entropy of a minimum mode as loss, and training the depth state space model by using the training sample set; and analyzing and processing the online industrial process data by using the trained depth state space model to determine whether the industrial