A Reinforcement-Learning-Augmented Liquid-Fueled Reactor Network Model for Predicting Lean Blowout in Gas Turbine Combustors

2026-07-21

AI Summary

This study addresses the insufficient accuracy and computational efficiency in predicting lean blowout (LBO) in gas turbine combustors by proposing a reinforcement learning–driven, multi-stage clustering–classification approach. The method first generates micro-clusters via k-means and then employs an Actor-Critic agent to merge them in a goal-oriented manner into optimal reaction zones, thereby constructing a reactor network model tailored for LBO prediction with liquid fuels. This work represents the first integration of goal-directed reinforcement learning into reactor network modeling, overcoming the limitations of conventional clustering strategies that rely on heuristic rules or spatial proximity. Coupled with a detailed Jet-A chemical mechanism (119 species and 841 reactions), the proposed framework captures LBO trends more accurately than standard k-means while significantly accelerating computation, enabling high-fidelity yet efficient exploration of design spaces.

Abstract

This study introduces a reinforcement learning (RL) framework for generating optimal liquid-fueled reactors to improve lean blowout (LBO) predictions in gas turbine combustors. Existing approaches for determining cluster boundaries rely on manual heuristics or distance-based metrics in the input space. In contrast, the proposed method is goal-oriented, explicitly accounting for the target metric (e.g., LBO prediction accuracy) during cluster formation. The framework employs a multi-stage clustering--classification strategy: an initial clustering step (e.g., $k$-means clustering) generates a large set of homogeneous micro-clusters, followed by an actor-critic RL agent that merges them into optimal reactor zones. The validation study, performed using a Jet-A mechanism (119 species, 841 reactions), shows the RL framework offers improved predictive fidelity compared to $k$-means and captures the correct LBO trends, while achieving substantial speedups relative to the high-fidelity computational model. Overall, the RL-driven approach demonstrates strong potential as a computationally efficient reduced-order modeling technique that can complement high-fidelity simulations for rapid design-space exploration.

Problem

• Lean Blowout
• Gas Turbine Combustors
• Liquid-Fueled Reactor
• Reduced-Order Modeling
• Prediction Accuracy

Innovation

• reinforcement learning
• lean blowout prediction
• reactor network model
• clustering optimization
• reduced-order modeling