A Learning-Based Decision Tool towards Smart Energy Optimization in the Manufacturing Process
Bibliographic record
Abstract
We developed a self-optimizing decision system that dynamically minimizes the overall energy consumption of an industrial process. Our model is based on a deep reinforcement learning (DRL) framework, adopting three reinforcement learning methods, namely: deep Q-network (DQN), proximal policy optimization (PPO), and advantage actor–critic (A2C) algorithms, combined with a self-predicting random forest model. This smart decision system is a physics-informed DRL that sets the key industrial input parameters to optimize energy consumption while ensuring the product quality based on desired output parameters. The system is self-improving and can increase its performances without further human assistance. We applied the approach to the process of heating tempered glass. Indeed, the identification and control of tempered glass parameters is a challenging task requiring expertise. In addition, optimizing energy consumption while dealing with this issue is of great value-added. The evaluation of the decision system under the three configurations has been performed and consequently, outcomes and conclusions have been explained in this paper. Our intelligent decision system provides an optimized set of parameters for the heating process within the acceptance limits while minimizing overall energy consumption. This work provides the necessary foundations to address energy optimization issues related to process parameterization from theory to practice and providing real industrial application; further research opens a new horizon towards intelligent and sustainable manufacturing.
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 | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".