NPDA fault classification method and control device based on multivariate discriminant analysis

• Main Authors: ZHANG HENGTAO, CHEN YAYONG, GAO SHENGJIE, ZHU YAOZONG, HOU CHAOJUN, LIU ZEFENG, CHENG ZHISHANG, ZHUANG JIAJUN, SUN SHENG, GUO QIWEI, MIAO AIMIN, TANG YU
• Format: Patent
• Language: Chinese; English
• Published: 08.11.2019
• Subjects: CONTROL OR REGULATING SYSTEMS IN GENERAL; CONTROLLING; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS ORELEMENTS; PHYSICS; REGULATING

The invention provides an industrial process fault classification method and device based on multivariate discriminant analysis. Firstly, based on a strategy of neighborhood preserving embedding algorithm local feature extraction, an optimal classification function taking a training data set, a dimensionality reduction dimension, a neighboring point and an Euclidean distance as input variables andtaking a dimensionality reduction transfer matrix as an output variable is established, so that while local features of input data and output data are preserved, neighboring point data is closer after dimensionality reduction in the same kind, and data in different kinds is more dispersive than the original; input and output features establishing data regression are obtained based on lower-dimension hidden variables of the data to establish a fault classification model based on NPDA. The industrial process fault classification method and device based on the multivariate discriminant analysissolve the problem that the