Xiaoqing Zheng, Bo Peng, Anke Xue, Ming Ge, Yaguang Kong, Aipeng Jiang
{"title":"Self-attention based Difference Long Short-Term Memory Network for Industrial Data-driven Modeling","authors":"Xiaoqing Zheng, Bo Peng, Anke Xue, Ming Ge, Yaguang Kong, Aipeng Jiang","doi":"10.1016/j.chemolab.2025.105535","DOIUrl":null,"url":null,"abstract":"<div><div>In modern industry, soft sensors provide real-time predictions of quality variables that are difficult to measure directly with physical sensors. However, in industrial processes, changes in material properties, catalyst deactivation, and other factors often lead to shifts in data distribution. Existing soft sensor models often overlook the impact of these distribution changes on performance. To address the issue of performance degradation due to changes in data distribution, this paper proposes a self-attention based Difference Long Short-Term Memory (SA-DLSTM) network for soft sensor modeling. By employing self-attention, industrial raw data is refined to facilitate the extraction of nonlinear features, thereby reducing the difficulty in modeling. A Difference Channel is designed to perform correlation analysis and select significant features from the raw data, followed by extracting the difference information that can reveal changes in the data distribution. The SA-DLSTM soft sensor model is established and validated on two benchmark industrial datasets: Debutanizer Column and Sulfur Recovery Unit. Comparisons with benchmark models, and state-of-the-art models show that SA-DLSTM achieves the best performance across all evaluation metrics, demonstrating the effectiveness of the proposed model.</div></div>","PeriodicalId":9774,"journal":{"name":"Chemometrics and Intelligent Laboratory Systems","volume":"267 ","pages":"Article 105535"},"PeriodicalIF":3.8000,"publicationDate":"2025-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Chemometrics and Intelligent Laboratory Systems","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0169743925002205","RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"AUTOMATION & CONTROL SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
In modern industry, soft sensors provide real-time predictions of quality variables that are difficult to measure directly with physical sensors. However, in industrial processes, changes in material properties, catalyst deactivation, and other factors often lead to shifts in data distribution. Existing soft sensor models often overlook the impact of these distribution changes on performance. To address the issue of performance degradation due to changes in data distribution, this paper proposes a self-attention based Difference Long Short-Term Memory (SA-DLSTM) network for soft sensor modeling. By employing self-attention, industrial raw data is refined to facilitate the extraction of nonlinear features, thereby reducing the difficulty in modeling. A Difference Channel is designed to perform correlation analysis and select significant features from the raw data, followed by extracting the difference information that can reveal changes in the data distribution. The SA-DLSTM soft sensor model is established and validated on two benchmark industrial datasets: Debutanizer Column and Sulfur Recovery Unit. Comparisons with benchmark models, and state-of-the-art models show that SA-DLSTM achieves the best performance across all evaluation metrics, demonstrating the effectiveness of the proposed model.
期刊介绍:
Chemometrics and Intelligent Laboratory Systems publishes original research papers, short communications, reviews, tutorials and Original Software Publications reporting on development of novel statistical, mathematical, or computer techniques in Chemistry and related disciplines.
Chemometrics is the chemical discipline that uses mathematical and statistical methods to design or select optimal procedures and experiments, and to provide maximum chemical information by analysing chemical data.
The journal deals with the following topics:
1) Development of new statistical, mathematical and chemometrical methods for Chemistry and related fields (Environmental Chemistry, Biochemistry, Toxicology, System Biology, -Omics, etc.)
2) Novel applications of chemometrics to all branches of Chemistry and related fields (typical domains of interest are: process data analysis, experimental design, data mining, signal processing, supervised modelling, decision making, robust statistics, mixture analysis, multivariate calibration etc.) Routine applications of established chemometrical techniques will not be considered.
3) Development of new software that provides novel tools or truly advances the use of chemometrical methods.
4) Well characterized data sets to test performance for the new methods and software.
The journal complies with International Committee of Medical Journal Editors'' Uniform requirements for manuscripts.