.png){kind=icon}
.png){kind=icon}
Return
Data-Driven Hierarchical Decision-Making Modeling for Complex Industrial Processes
.png){kind=icon}
PRE
AI
DOI：10.1109/TASE.2026.3677252.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Accurate decision-making for complex industrial processes is essential for improving resource utilization and optimizing operating conditions. However, current decision workflows operate within a hierarchical closed loop between the operations layer and the process-control layer. Cross-layer transmission of objectives and constraints and feedback coupling can degrade performance; the effect is pronounced under nonstationary conditions or when objectives or constraints change. To address these challenges, we propose the Bi-level Evolving Neural Networks (BENNs) framework with four components: 1) bi-level neural architecture that explicitly models cross-layer information flow and constraint propagation; 2) graph neural network-based structural similarity assessment that reduces redundant evaluations to improve search quality; 3) online evolving network strategy that adapts rapidly to process dynamics without extensive retraining; and 4) convex-hull knee-point-based compromise selection that encodes engineering pReferences. Experiments on real mineral processing data and on the Tennessee Eastman Process (TEP) show improvements over baselines in prediction accuracy, performance retention, and adaptation speed, demonstrating the potential of BENNs for complex industrial settings. Note to Practitioners—This data-driven hierarchical decision-making framework offers significant potential for practitioners in industries with complex, dynamic processes. Its application can streamline operations in areas such as mineral processing, chemical manufacturing, and wastewater treatment, where real-time adaptability is crucial. By using BENNs, plant operators can achieve enhanced control and optimization, responding effectively to fluctuating conditions. The framework’s ability to integrate human-like preferences enables decision-making to better align with specific operational goals. Furthermore, the improved optimization quality and continuous adaptation capabilities reduce downtime and improve overall efficiency. This approach empowers practitioners to move beyond rigid, expert-based systems, enabling proactive and intelligent control in evolving industrial environments.
Keywords:
Bi-level human decision-making
hierarchical neural network
evolutionary computation
industrial process
multi-population optimization
Journal
IF:
6.4
Papers:
4.9K
Citations:
1.6W