Industrial processes generate complex, interdependent time-series data from multiple sensors across various stages. Effective monitoring and timely anomaly detection of these time series through multivariate time series anomaly detection (MTAD) are crucial for preventing failures and ensuring the reliability of automated systems.
Graph neural networks (GNNs) have advanced MTAD by utilizing data-driven graphs to model complex dependencies among variables, effectively capturing relational structures within multivariate time series to enhance anomaly detection performance. However, existing GNN-based approaches often overlook critical process knowledge, and incorporating this knowledge into existing models remains challenging, leading to suboptimal performance.
To address this limitation, we propose a knowledge-assisted multi-graph framework for modeling sensor dependencies in multi-stage industrial processes for MTAD. This framework explicitly incorporates process knowledge into graph learning to enhance dependency modeling and improve anomaly detection performance. We construct three complementary graphs: one purely data-driven and two refined by integrating structural constraints derived from process knowledge.
To effectively leverage these graphs for anomaly detection, we employ a multi-graph attention network, enabling a more accurate and robust representation of complex dependencies. Comprehensive experiments on two real-world, multi-stage industrial datasets demonstrate that incorporating process knowledge substantially enhances anomaly detection performance.
Blogger's Review: This study successfully enhances the performance of graph neural networks in multivariate time series anomaly detection by integrating process knowledge, highlighting the importance of interdisciplinary knowledge in data-driven models. Future work should focus on optimizing this framework for more complex industrial environments.