One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case Study
Modern industrial mining and mineral processing applications are characterized by large volumes of historical process data. Hazardous events occurring in these processes compromise process safety and therefore overall viability. These events are recorded in historical data and are often preceded by...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-10-01
|
| Series: | Minerals |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-163X/11/10/1106 |
| _version_ | 1797513785618989056 |
|---|---|
| author | Carl Daniel Theunissen Steven Martin Bradshaw Lidia Auret Tobias Muller Louw |
| author_facet | Carl Daniel Theunissen Steven Martin Bradshaw Lidia Auret Tobias Muller Louw |
| author_sort | Carl Daniel Theunissen |
| collection | DOAJ |
| description | Modern industrial mining and mineral processing applications are characterized by large volumes of historical process data. Hazardous events occurring in these processes compromise process safety and therefore overall viability. These events are recorded in historical data and are often preceded by characteristic patterns. Reconstruction-based data-driven models are trained to reconstruct the characteristic patterns of hazardous event-preceding process data with minimal residuals, facilitating effective event prediction based on reconstruction residuals. This investigation evaluated one-dimensional convolutional auto-encoders as reconstruction-based data-driven models for predicting positive pressure events in industrial furnaces. A simple furnace model was used to generate dynamic multivariate process data with simulated positive pressure events to use as a case study. A one-dimensional convolutional auto-encoder was trained as a reconstruction-based model to recognize the data preceding the hazardous events, and its performance was evaluated by comparing it to a fully-connected auto-encoder as well as a principal component analysis reconstruction model. This investigation found that one-dimensional convolutional auto-encoders recognized event-preceding patterns with lower detection delays, higher specificities, and lower missed alarm rates, suggesting that the one-dimensional convolutional auto-encoder layout is superior to the fully connected auto-encoder layout for use as a reconstruction-based event prediction model. This investigation also found that the nonlinear auto-encoder models outperformed the linear principal component model investigated. While the one-dimensional auto-encoder was evaluated comparatively on a simulated furnace case study, the methodology used in this evaluation can be applied to industrial furnaces and other mineral processing applications. Further investigation using industrial data will allow for a view of the convolutional auto-encoder’s absolute performance as a reconstruction-based hazardous event prediction model. |
| first_indexed | 2024-03-10T06:21:27Z |
| format | Article |
| id | doaj.art-c2a7694a0c8f4aa8b25b81cb5ae715f0 |
| institution | Directory Open Access Journal |
| issn | 2075-163X |
| language | English |
| last_indexed | 2024-03-10T06:21:27Z |
| publishDate | 2021-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Minerals |
| spelling | doaj.art-c2a7694a0c8f4aa8b25b81cb5ae715f02023-11-22T19:16:57ZengMDPI AGMinerals2075-163X2021-10-011110110610.3390/min11101106One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case StudyCarl Daniel Theunissen0Steven Martin Bradshaw1Lidia Auret2Tobias Muller Louw3Department of Process Engineering, Faculty of Engineering, University of Stellenbosch, Stellenbosch 7600, South AfricaDepartment of Process Engineering, Faculty of Engineering, University of Stellenbosch, Stellenbosch 7600, South AfricaDepartment of Process Engineering, Faculty of Engineering, University of Stellenbosch, Stellenbosch 7600, South AfricaDepartment of Process Engineering, Faculty of Engineering, University of Stellenbosch, Stellenbosch 7600, South AfricaModern industrial mining and mineral processing applications are characterized by large volumes of historical process data. Hazardous events occurring in these processes compromise process safety and therefore overall viability. These events are recorded in historical data and are often preceded by characteristic patterns. Reconstruction-based data-driven models are trained to reconstruct the characteristic patterns of hazardous event-preceding process data with minimal residuals, facilitating effective event prediction based on reconstruction residuals. This investigation evaluated one-dimensional convolutional auto-encoders as reconstruction-based data-driven models for predicting positive pressure events in industrial furnaces. A simple furnace model was used to generate dynamic multivariate process data with simulated positive pressure events to use as a case study. A one-dimensional convolutional auto-encoder was trained as a reconstruction-based model to recognize the data preceding the hazardous events, and its performance was evaluated by comparing it to a fully-connected auto-encoder as well as a principal component analysis reconstruction model. This investigation found that one-dimensional convolutional auto-encoders recognized event-preceding patterns with lower detection delays, higher specificities, and lower missed alarm rates, suggesting that the one-dimensional convolutional auto-encoder layout is superior to the fully connected auto-encoder layout for use as a reconstruction-based event prediction model. This investigation also found that the nonlinear auto-encoder models outperformed the linear principal component model investigated. While the one-dimensional auto-encoder was evaluated comparatively on a simulated furnace case study, the methodology used in this evaluation can be applied to industrial furnaces and other mineral processing applications. Further investigation using industrial data will allow for a view of the convolutional auto-encoder’s absolute performance as a reconstruction-based hazardous event prediction model.https://www.mdpi.com/2075-163X/11/10/1106process monitoringfailure predictionsemi-supervised modelone-dimensional convolutional networkreconstruction-based model |
| spellingShingle | Carl Daniel Theunissen Steven Martin Bradshaw Lidia Auret Tobias Muller Louw One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case Study Minerals process monitoring failure prediction semi-supervised model one-dimensional convolutional network reconstruction-based model |
| title | One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case Study |
| title_full | One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case Study |
| title_fullStr | One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case Study |
| title_full_unstemmed | One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case Study |
| title_short | One-Dimensional Convolutional Auto-Encoder for Predicting Furnace Blowback Events from Multivariate Time Series Process Data—A Case Study |
| title_sort | one dimensional convolutional auto encoder for predicting furnace blowback events from multivariate time series process data a case study |
| topic | process monitoring failure prediction semi-supervised model one-dimensional convolutional network reconstruction-based model |
| url | https://www.mdpi.com/2075-163X/11/10/1106 |
| work_keys_str_mv | AT carldanieltheunissen onedimensionalconvolutionalautoencoderforpredictingfurnaceblowbackeventsfrommultivariatetimeseriesprocessdataacasestudy AT stevenmartinbradshaw onedimensionalconvolutionalautoencoderforpredictingfurnaceblowbackeventsfrommultivariatetimeseriesprocessdataacasestudy AT lidiaauret onedimensionalconvolutionalautoencoderforpredictingfurnaceblowbackeventsfrommultivariatetimeseriesprocessdataacasestudy AT tobiasmullerlouw onedimensionalconvolutionalautoencoderforpredictingfurnaceblowbackeventsfrommultivariatetimeseriesprocessdataacasestudy |