Entropy-enhanced batch sampling and conformal learning in VGAE for physics-informed causal discovery and fault diagnosis

Industry 4.0 has increased the demand for advanced fault detection and diagnosis (FDD) in complex industrial processes. This research introduces a novel approach to causal discovery and FDD using Variational Graph Autoencoders (VGAEs) enhanced with physics-informed constraints and conformal learning...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computers & chemical engineering Jg. 197; S. 109053
Hauptverfasser: Modirrousta, Mohammadhossein, Memarian, Alireza, Huang, Biao
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier Ltd 01.06.2025
Schlagworte:
Causal discovery
Conformal learning
Fault detection and diagnosis
Physics-informed constraints
Variational Graph Autoencoders
Variational Graph Autoencoders
Physics-informed constraints
Conformal learning
Fault detection and diagnosis
Causal discovery
ISSN:0098-1354
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Industry 4.0 has increased the demand for advanced fault detection and diagnosis (FDD) in complex industrial processes. This research introduces a novel approach to causal discovery and FDD using Variational Graph Autoencoders (VGAEs) enhanced with physics-informed constraints and conformal learning. Our method addresses limitations in conventional techniques, such as Granger causality, which struggle with high-dimensional, nonlinear systems. By integrating Graph Convolutional Networks (GCNs) and an entropy-based dynamic edge sampling method, the framework focuses on high-uncertainty regions of the causal graph. Conformal learning establishes rigorous thresholds for causal inference. Validated through simulation and case studies, including an Australian refinery and the Tennessee Eastman Process, our approach improves causal discovery accuracy, reduces spurious connections, and enhances fault classification. Integrating domain-specific physics information also led to faster convergence and reduced computational demands. This research provides an efficient, statistically robust approach for causal discovery and FDD in complex industrial systems. •Novel VGAE framework combines physics-informed constraints & conformal learning for causal discovery.•Dynamic edge sampling with entropy-based method targets high-uncertainty regions in process graphs.•Physics domain knowledge integration enhances model convergence and computational performance.•Framework validated through industry cases: Australian refinery & Tennessee Eastman process.•Method shows higher causal accuracy & fewer spurious links in complex industrial systems.
ISSN:0098-1354
DOI:10.1016/j.compchemeng.2025.109053