PLC operation behavior prediction method and device based on manifold learning, equipment and medium

Abstract

The invention provides a manifold learning-based PLC operation behavior prediction method and device, equipment and a medium, and the method comprises the steps: periodically collecting the operation variables of a PLC at a historical moment, forming an original state sequence, carrying out the preprocessing and continuous coding of the original state sequence, and generating a high-dimensional continuous state vector sequence; based on the high-dimensional continuous state vector sequence, constructing a nonlinear manifold and performing dimension reduction processing, and determining a low-dimensional embedding sequence and dimension reduction structure parameters; constructing a key variable mapping table, and carrying out serialized compression on the key variable mapping table, the high-dimensional continuous state vector sequence and the dimension reduction structure parameters to generate a file of an edge deployment model; and inputting the current operation behavior state into the edge deployment model for operation behavior prediction processing, and determining the operation behavior prediction state of the PLC at the next moment. The efficient, accurate and real-time prediction of the PLC operation behavior on the edge equipment is realized.

Images