Fault monitoring method based on multi-production unit variable cross-correlation decoupling strategy

Abstract

The invention discloses a fault monitoring method based on a cross-correlation decoupling strategy of multi-production unit variables, and aims to consider the cross-relationship between different production units in an industrial object into a distributed modeling and monitoring process from a data point of view, thereby implementing more reliable and efficient distributed fault monitoring. Specifically, firstly, according to the attribution of the measurement variables of each production unit, all the measured variables are divided into multiple variable sub-blocks. Secondly, the cross-correlation information between the sub-blocks of the variables and other sub-blocks of each variable is mined out by using the regression model. Finally, modeling and fault monitoring are implemented using errors after cross-correlation decoupling. Compared with the traditional method, the method utilizes a regression model to take into account the intersection relationship between different production unit variable sub-blocks, and monitors the error that can reflect whether the cross-correlation relationship between different production units changes, and therefore the method has more superior fault monitoring performance.

Description

technical field [0001] The invention relates to a data-driven fault monitoring method, in particular to a fault monitoring method based on a multi-production unit variable cross-correlation decoupling strategy. Background technique [0002] Ensuring continuous and normal production status is of great significance for reducing production costs and ensuring production safety. The technical means usually adopted are nothing more than real-time monitoring of process operation status, so as to identify abnormal status of the system in a timely manner. In recent years, with the advancement of large-scale industrial and "big data" construction, a large amount of real-time data can be collected in the production process, but an accurate mechanism model cannot be established. bedding. In this field of research, multivariate statistical process monitoring has received the most research and attention. Among them, the Principal Component Analysis (PCA) algorithm should be the most mai...

AI Technical Summary

Problems solved by technology

This is not advisable in the actual process, and it also violates the idea of ​​data-driven fault monitoring method from the perspective of data

Images