Stacking Ensemble Learning-Assisted Simulation of Plasma-Catalyzed CO2 Reforming of Methane
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have reported on the stacking ensemble learning-assisted simulation of plasma-catalyzed CO2 reforming of methane
I think the result of the manuscript is interesting for simulation approach, however I have few questions about the practical challenges on the plasma-catalyzed CO2 reforming of methane. I have the following comments on the manuscript.
- In the plasma-catalyzed CO2 reforming of methane, is there formation radical using combination CO2 and H2 gas?
- What does it mean by excited species? Is not in radical form?
- I think producing of methane in plasma-catalyzed CO2 reforming will be difficult as it plasma will radically form plasma ions.
- Authors should provide a table to understand of the present status of the work.
- What kind plasma generation process can be used to catalyzed CO2 reforming of methane, such as, microwave plasma, arc discharge plasma etc.
Based on the above comments, I recommend major revision of the manuscript for further consideration of publication in Electronics.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsAbstract
Line 6–8: “The machine learning model trained on limited data... even at non-convergence points.” → Clarify how the ML model handles non-convergence points. Specify if it extrapolates or interpolates.
Line 13–14: “integrating principles and data, and is an effective research method in the field of energy conversion.” → Generic phrasing. Replace with concrete outcomes (e.g., “reducing computational time by X% compared to traditional methods”).
Line 21: “Plasma-catalysis technology, in particular, has demonstrated significant potential...” → Overly generic. Specify unique advantages (e.g., activation pathways, selectivity).
Introduction
Line 22–23: “Plasma-catalysis technology... for COâ‚‚ conversion.” → Overly broad. Specify which conversion processes (e.g., methane reforming, ammonia synthesis).
Line 39–41: “Zhong et al. have proposed a deep learning method... transient arcs with radial velocity.” → Clarify novelty compared to prior work (e.g., Mathews et al.).
Line 54: “machine learning techniques have demonstrated unique advantages...” → Repetitive phrasing. Replace with specific advantages (e.g., handling non-linearities, scalability).
Methodology
Section 2.1:
Equation 1: Define all variables (e.g., αijLeft\alpha^{Left}_{ij}αijLeft​, αijRight\alpha^{Right}_{ij}αijRight​) in the text. Equations 2–5: Clarify if vvv (thermal velocity in Eq. 2 vs. surface diffusion hopping frequency in Eq. 4) refers to the same parameter. Use distinct notation if different.
Line 110: “v also represents the surface diffusion hopping frequency” → Confusing dual definition of v. Clarify distinct variables.
Line 196: “This approach effectively resolves issues of non-convergence...” → Vague. Replace with concrete examples of resolved convergence scenarios.
Section 2.2: Missing details on cross-validation implementation. Specify how training/test splits were managed and how meta-model training avoids data leakage.
Results
Section 3.1: Table 2: Title incorrectly repeats “Species contained in... methane.” Replace with “Optimized hyperparameters for base models.”
Line 341–344: “Parts a-c... RMSE values.” → Figures 3a–f are referenced but missing. Confirm inclusion in submission.
Line 385: “predictions of the stacking ensemble model are significantly closer to the true values” → Clarify if “true values” are experimental or simulation-derived. If the latter, note limitations.
Table 2: XGBoost “max depth=45” is unusually high. Justify or address overfitting risks (e.g., regularization parameters).
Discussion & Conclusions
Line 444: “significantly improves simulation efficiency” → Quantify efficiency gains (e.g., runtime comparison with traditional methods).
Line 446–449: “Future advancements... industrial scale.” → Non-specific. Propose concrete research directions (e.g., real-time optimization, multi-scale coupling).
No comparison with physical experiments was conducted. Include experimental CO density data to validate simulations.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript is revised according to the comments of the reviewer. I would like to recommend acceptance of the manuscript in present form.
Author Response
Dear Reviewer:
Thank you for your professional review and valuable time on our manuscript [electronics-3506082/title "Stacking Ensemble Learning Assisted Simulation of Plasma Catalyzed CO2 Reforming of Methane"]. We are pleased to learn that you believe the manuscript can be accepted without any modifications, which is a great encouragement to our team.
Thank you again for your support and assistance!
Yours sincerely,
Shaohua Qin
E-mail: qinshaohua@sdnu.edu.cn
Reviewer 2 Report
Comments and Suggestions for AuthorsAll my comments were addressed, except:
Comment 14: the author should incorporate the response into the manuscript and discuss it in more depth.
Comment 17: the author should conduct a (brief) discussion in the Results section.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
Comments and Suggestions for AuthorsAll my comments were addressed.