Intermittent process online modal identification method based on double-boundary SVDD

The invention discloses an intermittent process online modal recognition method based on double-boundary SVDD, and the method comprises the steps: firstly standardizing the historical batch data of an intermittent process, carrying out the modal division through time sequence fuzzy C-means clustering, and obtaining the data of each stable modal and transition modal; then training a DBSVDD model by using data of each stable mode in the intermittent process to obtain a sphere center and an inner boundary of a hypersphere, calculating an outer boundary of the DBSVDD model by using data of each stable mode and adjacent transition modes, and constructing a double-boundary SVDD online mode recognition model; and finally, realizing intermittent process online modal identification by utilizing the inner and outer boundaries of the double-boundary SVDD model and an online modal identification strategy. According to the method, the time sequence characteristics of the intermittent process data and the time continuity o