To address the challenges of strong variable coupling, multi-scale feature extraction, and poor robustness to missing data in industrial process quality prediction, this paper proposes a Dilated Multi-scale Convolutional Gated Network (DMCGN) for data-driven soft sensors. The DMCGN integrates dilated convolution layers (dilation rates r = 1, 2, 3) to expand the receptive field, multi-scale convolutional kernels (3 × 3, 5 × 5, 7 × 7) to capture local spatio-temporal features, and a gated aggregation module to adaptively suppress redundant information. Validations on two real industrial datasets (sulfur recovery unit, SRU; debutanizer column process, DCP) show that the proposed method outperforms state-of-the-art models: for SRU, the RMSE is 0.0095 and R2 reaches 0.9731; for DCP, the RMSE is 0.0081 and R2 is 0.9984. Additionally, the DMCGN maintains high prediction accuracy under 10%–30% data missing rates, providing a reliable solution for real-time quality monitoring in complex industrial processes.
LiZJinHDongS, et al. Semi-supervised ensemble support vector regression based soft sensor for key quality variable estimation of nonlinear industrial processes with limited labeled data. Chem Eng Res Des2022; 179: 510–526.
2.
WeiBTanSZhangQ, et al. A hybrid soft sensor for key product yield of FCC unit based on deep learning framework driven by data and process mechanism. Chem Eng Res Des2024; 202: 429–443.
3.
LawrenceNPDamarlaSKKimJW, et al. Machine learning for industrial sensing and control: a survey and practical perspective. Control Eng Pract2024; 145: 105841.
4.
Acosta-PavasJCRobles-RodriguezCEGriolD, et al. Soft sensors based on interpretable learners for industrial-scale fed-batch fermentation: learning from simulations. Comput Chem Eng2024; 187: 108736.
5.
SunQGeZ. A survey on deep learning for data-driven soft sensors. IEEE Trans Ind Inform2021; 17: 5853–5866.
6.
JiangYYinSDongJ, et al. A review on soft sensors for monitoring, control, and optimization of industrial processes. IEEE Sens J2021; 21: 12868–12881.
7.
KayHKaySMowbrayM, et al. Constructing a symbolic regression-based interpretable soft sensor for industrial data analytics and product quality control. Ind Eng Chem Res2024; 63: 4083–4092.
8.
SadeghianAMagbool JanNWuO, et al. Robust probabilistic principal component regression with switching mixture Gaussian noise for soft sensing. Chemometr Intell Lab Syst2022; 222: 104491.
9.
JiangQYanXYiH, et al. Data-driven batch-end quality modeling and monitoring based on optimized sparse partial least squares. IEEE Trans Ind Electron2020; 67: 4098–4107.
10.
DilmiS. Calcium soft sensor based on the combination of support vector regression and 1-D digital filter for water quality monitoring. Arab J Sci Eng2023; 48: 6111–6136.
11.
ChitralekhaSBShahSL. Application of support vector regression for developing soft sensors for nonlinear processes. Can J Chem Eng2010; 88: 696–709.
12.
GonzagaJCBMeleiroLACKiangC, et al. ANN-based soft-sensor for real-time process monitoring and control of an industrial polymerization process. Comput Chem Eng2009; 33: 43–49.
13.
TercanHMeisenT. Machine learning and deep learning based predictive quality in manufacturing: a systematic review. J Intell Manuf2022; 33: 1879–1905.
14.
SarkerIH. Deep learning: a comprehensive overview on techniques, taxonomy, applications and research directions. SN Comput Sci2021; 2: 420.
15.
ShangCYangFHuangD, et al. Data-driven soft sensor development based on deep learning technique. J Process Control2014; 24: 223–233.
16.
RenLWangTLailiY, et al. A data-driven self-supervised LSTM-DeepFM model for industrial soft sensor. IEEE Trans Ind Inform2022; 18: 5859–5869.
17.
GengZChenZMengQ, et al. Novel transformer based on gated convolutional neural network for dynamic soft sensor modeling of industrial processes. IEEE Trans Ind Inform2022; 18: 1521–1529.
18.
YangBLuX. Online prediction of joint mechanical properties of FSW thick AA2219-T8 based on multi-source information fusion using 1DCNN. Proc Inst Mech Eng B J Eng Manuf2025; 239: 1491–1508.
19.
YuFKoltunV. Multi-scale context aggregation by dilated convolutions. In: International Conference on Learning Representations (ICLR), 2016.
YinYTianJLiuX. Remaining useful life prediction based on parallel multi-scale feature fusion network. J Intell Manuf2025; 36: 3111–3127. https://doi.org/10.1007/s10845-024-02399-y
22.
ZhuQXiongQYangZ, et al. A novel feature-fusion-based end-to-end approach for remaining useful life prediction. J Intell Manuf2023; 34: 3495–3505.
23.
WangCWangHLiuM. A dual-attention feature fusion network for imbalanced fault diagnosis with two-stream hybrid generated data. J Intell Manuf2024; 35: 1707–1719.
24.
ChenSHuangZWangT, et al. Wafer map defect recognition based on multi-scale feature fusion and attention spatial pyramid pooling. J Intell Manuf2025; 36: 271–284. https://doi.org/10.1007/s10845-023-02231-z
25.
RatoTJReisMS. Multiresolution soft sensors: a new class of model structures for handling multiresolution data. Ind Eng Chem Res2017; 56: 3640–3654.
26.
BakshiBRStephanopoulosG. Wave-net: a multiresolution, hierarchical neural network with localized learning. AIChE J1993; 39: 57–81.
27.
LiuKLuNWuF, et al. Model fusion and multiscale feature learning for fault diagnosis of industrial processes. IEEE Trans Cybern2023; 53: 6465–6478.
28.
LiuYDillonTYuW, et al. Missing value imputation for industrial IoT sensor data with large gaps. IEEE Internet Things J2020; 7: 6855–6867.
29.
DuWCôtéDLiuY. SAITS: self-attention-based imputation for time series. Expert Syst Appl2023; 219: 119619.
30.
YuanXHuangLYeL, et al. Quality prediction modeling for industrial processes using multiscale attention-based convolutional neural network. IEEE Trans Cybern2024; 54: 2696–2707.
31.
WangPChenPYuanY, et al. Understanding convolution for semantic segmentation. In: 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), 2018, pp. 1451–1460.
32.
ChenLCPapandreouGKokkinosI, et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Trans Pattern Anal Mach Intell2018; 40: 834–848.
PatanèLXibiliaMG. Echo-state networks for soft sensor design in an SRU process. Inf Sci2021; 566: 195–214.
35.
FortunaLRizzoASinatraM, et al. Soft analyzers for a sulfur recovery unit. Control Eng Pract2003; 11: 1491–1500.
36.
ZhaoYDengXLiS. A nonlinear industrial soft sensor modeling method based on locality preserving stochastic configuration network with utilizing unlabeled samples. ISA Trans2023; 139: 548–560.
37.
FortunaLGrazianiSXibiliaMG. Soft sensors for product quality monitoring in debutanizer distillation columns. Control Eng Pract2005; 13: 499–508.
38.
YuanXLiLWangY, et al. Deep learning for quality prediction of nonlinear dynamic processes with variable attention-based long short-term memory network. Can J Chem Eng2020; 98: 1377–1389.
39.
HochreiterSSchmidhuberJ. Long short-term memory. Neural Comput1997; 9: 1735–1780.
40.
ChenXZhangBGaoD. Bearing fault diagnosis base on multi-scale CNN and LSTM model. J Intell Manuf2021; 32: 971–987.
41.
YangKWangYTangY, et al. A temporal convolution and gated recurrent unit network with attention for state of charge estimation of lithium-ion batteries. J Energy Storage2023; 72: 108774.
42.
WangSShiJYangW, et al. High and low frequency wind power prediction based on transformer and BiGRU-attention. Energy2024; 288: 129753.
