Intermittent process fault detection method and system considering two-dimensional dynamic characteristics

Abstract

The invention provides an intermittent process fault detection method and system considering two-dimensional dynamic characteristics. The method comprises the following steps: extracting low-dimensional characteristic information of test data by utilizing a load matrix of a two-dimensional dynamic kernel slow characteristic analysis model, and calculating monitoring statistics of the test data ina principal component space and a residual error space; judging whether a fault occurs in the intermittent process or not according to a comparison result of the monitoring statistics and the corresponding control limit, wherein each batch data set in the three-dimensional training data set is expanded by utilizing an autoregressive moving average time sequence model to obtain a corresponding augmented batch data set, and the training data is normal operation condition data of the intermittent process; nonlinearly mapping the augmented batch data set to a high-dimensional feature space, establishing a time dynamic kernel slow feature analysis model, introducing a kernel function skill to calculate a kernel matrix and a time change kernel matrix, calculating a total average kernel matrix and a total time change kernel matrix based on a global modeling strategy, and constructing a two-dimensional dynamic kernel slow feature analysis model.

Description

Technical field[0001]The invention belongs to the technical field of dynamic nonlinear multivariable intermittent process fault detection, and in particular relates to an intermittent process fault detection method and system considering two-dimensional dynamic characteristics.Background technique[0002]The statements in this section merely provide background information related to the present invention, and do not necessarily constitute prior art.[0003]As the batch process is increasingly highly integrated, large-scale and complex, the fault detection of the batch process has become a key technology to ensure its safe and stable operation. With the development of modern computer control technology, a wealth of process operation data is collected and stored during intermittent processes. Therefore, data-driven fault detection technology has gradually become a research hotspot in the field of intermittent process monitoring. Researchers have proposed a series of data-driven fault dete...

AI Technical Summary

Problems solved by technology

Although the slow feature analysis technology has achieved certain application results in the field of fault detection of batch processes, the inventors found that the shortcomings of the slow feature analysis technology in the field of fault detection of batch processes are: (1) the batch process has a two-dimensional nature in nature Dynamic characteristics: The dynamic characteristics of the batch dimension and the dynamic characteristics of the time dimension. Although the slow feature analysis can deal with the dynamic characteristics of the batch process in the time dimension, it cannot eliminate the dynamic characteristics of the batch dimension of the batch process, which will affect the effect of fault detection.

(2) Slow feature analysis is actually a linear dimensionality reduction method, which cannot deal with the strong nonlinear characteristics of intermittent processes and reduces the performance of fault detection

Images