Data Mode Related Interpretable Transformer Network for Predictive Modeling and Key Sample Analysis in Industrial Processes

Liu, Diju; Wang, Yalin; Liu, Chenliang; Yuan, Xiaofeng; Yang, Chunhua; Gui, Weihua

doi:10.1109/tii.2022.3227731

Cited by 34 publications

(5 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the one hand, non-contact measurement methods such as digital image correlation (DIC) will be combined to conduct research on the sensing accuracy of the system based on specific measurement objects and other factors such as the installation method of WRFBGs. On the other hand, the detection technology proposed in this paper for model validation will be introduced to achieve accurate prediction of difficult to measure information, such as modeling ideas based on deep learning [34][35].…”

Section: Discussionmentioning

confidence: 99%

Real-Time Monitoring of Strain Processes With Large-Range and High-Spatial Resolution Using the Method of Weak Reflection FBG Measurement Based on OFDR

Chen,

Li,

Yang

et al. 2024

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

The real-time monitoring of strain processes that can simultaneously realize a wide range and high spatial resolution has always been a huge challenge and cannot be satisfied by traditional electrical and non-contact measurement methods. In this article, the OFDR collocated with weak reflection FBG sensor (WRFBGs) was proposed to actively carry out multiple experiments on the non-standard reinforcement samples with complex strain processes in order to overcome this challenge. According to the result analysis, the complex strain variation law of the surface of the sample and the large span strain of adjacent locations during the whole process of tensile can be obtained by this method. The yield point (YP) can be accurately captured and positioned with mm-level horizontal precision. At the same time, real-time monitoring can be realized for the step process when the strain of the point increases to more than 25000με instantly. The method can also realize the strain detection capability with a response speed of 38 Hz and a 1.28 mm high spatial resolution with monitoring the real-time high-density and large strain distribution.

show abstract

Section: Discussionmentioning

confidence: 99%

Real-Time Monitoring of Strain Processes With Large-Range and High-Spatial Resolution Using the Method of Weak Reflection FBG Measurement Based on OFDR

Chen,

Li,

Yang

et al. 2024

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

show abstract

“…In order to achieve the goal of intelligent ironmaking, a growing number of researchers have started to establish a comprehensive industrial automatic system, including key performance indicators (KPIs) prediction, optimization, control, and decision-making. Accurate and efficient KPIs prediction, as an important prerequisite for industrial automation systems, has received widespread attention from both academic and industry fields. − …”

Section: Introductionmentioning

confidence: 99%

“…Accurate and efficient KPIs prediction, as an important prerequisite for industrial automation systems, has received widespread attention from both academic and industry fields. 2 − 6 …”

Section: Introductionmentioning

confidence: 99%

A Survey of Data-Driven Soft Sensing in Ironmaking System: Research Status and Opportunities

Yan,

Kong,

et al. 2024

ACS Omega

View full text Add to dashboard Cite

Data-driven soft sensing modeling is becoming a powerful tool in the ironmaking process due to the rapid development of machine learning and data mining. Although various soft sensing techniques have been successfully used in both the sintering process and blast furnace, they have not been comprehensively reviewed. In this work, we provide an overview of recent advances on soft sensing in the ironmaking process, with a special focus on data-driven techniques. First, we present a general soft sensing development framework of the ironmaking process based on the mechanism analysis and process characteristics. Second, we provide a detailed taxonomy of current soft sensing methods categorized by their predictive tasks (i.e., quality indicators prediction, state parameters prediction, etc.). Finally, we outline several insightful and promising directions, such as self-supervised learning and digital twins in the ironmaking process, for future research.

show abstract

“…In recent years, extensive research has been conducted on datadriven soft sensor by many scholars. Liu et al [14] proposed a novel data modal correlation interpretable transformer network called DMRI-Former, designed for predictive modelling and critical sample analysis in industrial processes. This network aims to achieve accurate prediction of difficult-tomeasure quality variables in industrial process control and optimization.…”

Section: Introductionmentioning

confidence: 99%

Ensemble learning soft sensor method of endpoint carbon content and temperature of BOF based on GCN embedding supervised ensemble clustering

Gu,

Liu,

Chen

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

The endpoint control of Basic Oxygen Furnace (BOF) steelmaking depends on the prediction of the endpoint carbon content and temperature. However, predicting these variables is challenging because of the numerous working conditions in the industrial field and the volatility of the sensor data collected during BOF steelmaking. The accuracy of prediction models in ensemble learning depends significantly on the initial distribution of data. However, the complex nature of BOF steelmaking data makes it challenging to generate diverse subsets, which ultimately affects the accuracy of predictions. This paper presents a new approach called Graph Convolutional Network Node Embedding Supervised Ensemble Clustering (GESupEC) for soft sensor modelling in ensemble learning to tackle these issues. GESupEC utilizes a similarity graph derived from a co-association matrix and employs graph convolutional networks to extract structural information among nodes. By optimising the clustering loss within the network, GESupEC learns compact node representations that are useful for the clustering task. Furthermore, it generates a reconstruction matrix based on the similarity of node embeddings. This matrix helps with the extraction of a suitable subset of data for BOF steelmaking through matrix decomposition. After that, the gradient boosting decision tree regression sub-model is established based on the data subset. An ensemble strategy called Gray Relational Analysis Weighted Average is proposed, which assigns weights based on the grey relation similarity between test samples and different data subsets. This weighted average strategy aims to enhance the accuracy of carbon content and temperature predictions. When tested with actual BOF steelmaking generation process data, the prediction accuracy of carbon content reached 88.6% within the error range of ±0.02%, and the prediction accuracy of temperature reached 92.6% within the error range of ±10 °C.

show abstract

Data Mode Related Interpretable Transformer Network for Predictive Modeling and Key Sample Analysis in Industrial Processes

Cited by 34 publications

References 33 publications

Real-Time Monitoring of Strain Processes With Large-Range and High-Spatial Resolution Using the Method of Weak Reflection FBG Measurement Based on OFDR

Real-Time Monitoring of Strain Processes With Large-Range and High-Spatial Resolution Using the Method of Weak Reflection FBG Measurement Based on OFDR

A Survey of Data-Driven Soft Sensing in Ironmaking System: Research Status and Opportunities

Ensemble learning soft sensor method of endpoint carbon content and temperature of BOF based on GCN embedding supervised ensemble clustering

Contact Info

Product

Resources

About