EKG-AC: A New Paradigm for Process Industrial Optimization Based on Offline Reinforcement Learning With Expert Knowledge Guidance
Bibliographic record
Abstract
Operation optimization plays a crucial role in process control, directly influencing product quality and profitability. Reinforcement learning (RL), with its capabilities in autonomous learning and dynamic adaptability, has become a promising solution in this domain. However, its real-world application is constrained by the high costs and risks associated with its interactions with environments. Offline RL, which leverages fixed datasets without interactions, offers an alternative but faces significant challenges in the process industry due to imbalanced multioperating condition scenarios and heightened safety sensitivity. To address these challenges, this article introduces a novel offline actor-critic algorithm with expert knowledge guidance (EKG-AC). The method begins with a diffusion-transformer-based action generation framework that mitigates the out-of-distribution problem by capturing the evolution of decision sequences and the interdependencies between states and actions. An expert knowledge guidance mechanism is then integrated, steering the model to generate safe and adaptive candidate actions aligned with current operating conditions and expert knowledge. Subsequently, within the actor-critic framework, the optimal action is selected from the candidate pool based on the evaluated Q-value, thereby setting the operational variables for the optimization task. The proposed algorithm is validated through two real-world industrial processes, demonstrating superior optimization performance and behavior that is closely aligned with expert decision-making, underscoring its substantial practical value.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".