Intermittent process multi-mode division method for time sequence constraint fuzzy clustering

• Main Authors: HAN RUI, WANG JIANLIN, QIU KEPENG
• Format: Patent
• Language: Chinese; English
• Published: 14.05.2019
• Subjects: CALCULATING; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses an intermittent process multi-mode division method for time sequence constraint fuzzy clustering, and belongs to the technical field of intermittent process monitoring. The method comprises the following steps: expanding multiple batches of process data of an intermittent process into a two-dimensional data matrix according to a sampling time direction, standardizing the two-dimensional data matrix, and constructing a standardized process data sample set; secondly, dividing all process data samples of the intermittent process into the same class, and calculating the initial membership degree of each sample through a fuzzy clustering algorithm; and finally, starting from the initial sampling moment, adding the process data samples one by one as a new time sequence sample set, iteratively calculating the membership degree of the newly added samples, and dividing the modality of the intermittent process by comparing the membership degree of the newly added sampleswith the initial membersh