Process monitoring and fault diagnosis method based on tree structure sparsity

Abstract

The invention discloses a process monitoring and fault diagnosis method based on tree structure sparsity. Training data and test data are collected, and principal components of the training data and process monitoring statistic control limits are extracted through principal component analysis; process monitoring statistics of the test data are calculated, and a comparison is made and whether a fault occurs or not is judged; if a fault is monitored, a fault isolation model is established for processing a fault diagnosis problem, training data is converted into a tree structure, the weight of each node in the tree structure is established, and a fault diagnosis model based on tree structure sparsity is established according to the fault isolation model; and the fault diagnosis model based ontree structure sparseness is solved to obtain an optimal fault amplitude, and process variables of the fault are positioned and separated by utilizing the optimal fault amplitude. The method can be conveniently expanded to a parallel or distributed version, can meet the requirements for solving speed and precision of a large data set in the industrial process, and provides effective support for industrial production control behaviors.

Description

technical field [0001] The invention belongs to a monitoring method in the field of process monitoring and fault diagnosis in an industrial control system, and in particular relates to a process monitoring and fault diagnosis method based on tree structure sparseness. Background technique [0002] With the rapid development of sensing and instrumentation technology, the amount of data collected and analyzed in modern industrial factories is increasing exponentially. Abundant data information greatly facilitates the task of process monitoring, however, it also brings great challenges due to problems such as high dimensionality, multi-scale, inconsistent data quality, etc. To handle large data sets in large-scale processes, different distributed process monitoring methods, such as distributed principal component analysis (PCA) and distributed canonical correlation analysis (CCA), have been studied. The distributed process monitoring method decomposes the process into a group ...

AI Technical Summary

Problems solved by technology

However, it has been proven that for highly correlated process variables, the Lasso-like problem leads to a decrease in isolation accuracy

Images