Supervised deep learning algorithms for process fault detection and diagnosis under different temporal subsequence length of process data
Fault detection and diagnosis (FDD) play a vital role in abnormal situation management of chemical industrial processes. Current FDD technologies mostly rely on data-driven solutions by making full use of abundant process data collected by the state-of-the-art distributed process instruments and sen...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | en |
| Published: |
Springer Nature
2025
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/45078/1/FULLTEXT.pdf https://eprints.ums.edu.my/id/eprint/45078/ https://doi.org/10.1007/s10489-025-06711-y |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Fault detection and diagnosis (FDD) play a vital role in abnormal situation management of chemical industrial processes. Current FDD technologies mostly rely on data-driven solutions by making full use of abundant process data collected by the state-of-the-art distributed process instruments and sensors. Deep learning algorithms were widely used among all the data-driven algorithms. Industrial process time series data could be processed with ease by deep learning algorithms, particularly transformer-based models because of their multi-head attention mechanism. Different lengths of snippets of sequence (or subsequence) would have a multitude of perspectives viewed by the deep learning algorithms, subsequently impacting their FDD performance. This study, therefore, aims to investigate the effects of varying subsequence lengths on the FDD performance of common deep learning algorithms, consisting of a multilayer perceptron, convolutional neural network, long short-term memory, transformer, industrial process optimization—vision transformer (IPO-Vitt) using two benchmark case studies, namely continuous stirred tank reactor (CSTR) and Tennessee Eastman Process (TEP). Addition- ally, faulty data are rare to occur, and the fault labelling process is generally tedious and expensive to perform. The effects of labelled training data sizes were also studied on the FDD performance. The findings clearly indicate that the IPO-Vitt, a variant of transformer-based models, exhibited the best FDD performance under 10% and 50% subsequence length of data on CSTR and TEP case studies, respectively, for optimal feature extraction, even with 10% of fully labelled input data. |
|---|