Abstract:Data-driven modeling based on deep learning is crucial for online product quality prediction in industrial processes. Extracting latent data interactions from sensor variables lies at the core of various data-driven modeling applications. Typically, the observed variables exhibit non-stationary properties due to changes in operating conditions and sensor tuning issues. These fluctuations inevitably affect the reliability of traditional feature extraction methods, thus hindering their applications. Hence, this paper proposes a novel method named residual-aware deep attention graph convolutional network (RaDA-GCN) to explore potential interactions among sensor variables. RaDA-GCN ingeniously incorporates an attention mechanism into the graph convolution layer to extract nonlinear variable-related features according to their importance. Then, a novel residual-aware connection module is designed to reduce data uncertainty and alleviate over-smoothing. By skillfully stacking multiple attention map convolutional layers with the integration of residual-aware connections, deep structural features are obtained to facilitate the effective quantification and revelation of potential relationships among data variables. Finally, the application of the predictive modeling framework based on the proposed method is used to validate its effectiveness in actual industrial process data. The experimental results demonstrate that the proposed RaDA-GCN method exhibits a 23% improvement in the R-squared (R 2 ) indicator and a 13% reduction in the root mean square error (RMSE) compared to the traditional graph convolutional network method.

Hierarchical neural network based product quality prediction of industrial ethylene pyrolysis process

Direct Prediction of Steam Cracking Products from Naphtha Bulk Properties: Application of the Two Sub-Networks ANN

A Novel Modeling Approach and Its Application in Polymer Quality Index Prediction

Gated Quality-Driven Feature Extraction with A Hybrid SIAE Model for Industrial Soft Sensor

Multiple-prior-knowledge Neural Network for Industrial Processes

Soft-sensing research on the gas phase ethylene polymerization in fluidized bed reactor based on DPCA-RBF network

A Multiphase Dual Attention-Based LSTM Neural Network for Industrial Product Quality Prediction

Path Enhanced Bidirectional Graph Attention Network for Quality Prediction in Multistage Manufacturing Process

Multivariate Deep Reconstruction Neural Network for Multi-step-ahead Prediction of Industrial Process Quality Variables

A new approach for quality prediction and control of multistage production and manufacturing process based on Big Data analysis and Neural Networks

Neural Soft-Sensor of Product Quality Prediction

Pyramid Dynamic Bayesian Networks for Key Performance Indicator Prediction in Long Time-Delay Industrial Processes

Residual-aware deep attention graph convolutional network via unveiling data latent interactions for product quality prediction in industrial processes

An Adversarial Bidirectional Serial–parallel LSTM-based QTD Framework for Product Quality Prediction

Product quality prediction in dense medium coal preparation process based on recurrent neural network

Prediction of the Quality of Dairy Products Based on the BP Neural Network

An Innovative Graph Neural Network Model for Detailed Effluent Prediction in Steam Cracking

Deep Learning of Partially Labeled Data for Quality Prediction Based on Stacked Target-Related Laplacian Autoencoder.

Analyzing And Modeling Ethylene Cracking Process With Complex Networks Approach

Concurrent Analytics of Temporal Information and Local Correlation for Meticulous Quality Prediction of Industrial Processes

Prediction of Oil Quality Based on Neural Network