Intermittent process fault identification method and system based on TDKSFA nonlinear contribution graph

Abstract

The invention provides an intermittent process fault identification method and system based on a TDKSFA nonlinear contribution graph. The method comprises the following steps: performing mean centralization operation on pseudo sample kernel matrixes corresponding to a to-be-identified fault data set and an average training data set in sequence, and respectively projecting the pseudo sample kernelmatrixes to a load matrix of a TDKSFA model to obtain corresponding score matrixes; defining a nonlinear contribution degree of any process variable relative to the fault according to the difference between two pseudo sample projection tracks, and conducting calculating to obtain nonlinear contributions of all process variables relative to the fault; identifying the fault variable according to thenonlinear contribution of each process variable, and determining the fault type of the to-be-identified fault data; expanding the average training number set by utilizing an autoregressive moving average time sequence model, and calculating a corresponding augmented average training data set by adopting a global modeling strategy; and nonlinearly mapping the augmented average training data set toa high-dimensional feature space, and introducing a kernel function skill to calculate a kernel matrix and a time-varying kernel matrix to construct a TDKSFA model.

Description

technical field [0001] The invention belongs to the field of dynamic nonlinear multivariable intermittent process fault identification, in particular to a batch process fault identification method and system based on a TDKSFA nonlinear contribution graph. 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 tends to be highly integrated, large-scale and complex, the fault diagnosis 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 in batch processes. Therefore, data-driven fault diagnosis technology has gradually become a research hotspot in the field of batch process monitoring. Researchers have proposed a series of data-driven fault diagnosis methods, such as prin...

AI Technical Summary

Problems solved by technology

Although the method based on KSFA has achieved certain application results, the inventors found that the disadvantages of the KSFA method are: (1) the batch process has two-dimensional dynamic characteristics in essence: the dynamic characteristics of the batch dimension and the dynamic characteristics of the time dimension, However, the KSFA method cannot deal with the two-dimensional dynamic characteristics of the batch process at the same time, which reduces the performance of fault diagnosis.

(2) The KSFA method has been used to detect faults in nonlinear intermittent processes, but it is only used to detect faults, and the type of fault data to be identified cannot be identified after the fault is detected.

Images