Optimal Control Algorithm for Subway Train Operation by Proximal Policy Optimization
Round 1
Reviewer 1 Report
The manuscript reports the new optimization algorithm for the optimal control of subway train operation. Detailed simulation results of algorithms are presented based on the actual track and train conditions in Beijing. The manuscript reads well. The figures are clear and readable and support the findings. The research is interesting and my main concern is the structure of the Conclusion section which lacks the following justifications:
1. The novelty explanation is desirable, what is new in this work? Refer to your findings and the data.
2. Authors should critically point out their weak parts and discussed them to support some future research in this subject.
Also, the Reference list may be expanded by related topic papers published in MDPI journals.
This manuscript represents an
interesting research regarding the new optimization algorithm for the
optimal control of subway train operation. More specifically, the title
and abstract are appropriate for the content of the text. In abstract,
the authors summarized the main research question and key findings. The
Introduction section and Literature review are carefully done.
Furthermore, the manuscript is well constructed and analyses are well
performed. Data collected are interpreted accurately and the results
support the discussion. I do not have remarks on figures as they are
clear and readable and support the findings. The Reference list is
appropriate and the cited references are mostly recent publications, but
it could be expanded with related topic papers published in MDPI
journals.
My main remarks refer to Conclusion section as it is not satisfactory.
First, the authors should explain the topic and then the purpose of
their research. Then, to summarize their findings that emphasize the
importance and novelty of this research. The opportunities for future
research analysis should be indicated. Also, the limitations of the
present research methods are missing.
Kind regards,
Author Response
Thanks for your comments!
The topic and the purpose are added to this paper as well as the importance and novelty of this research. Also, we introduced the limitations of the present research and the future research analysis in the new version of this paper. And new preferences are cited in this paper.
Reviewer 2 Report
I recommend to complete the literature review with more current relevant sources with the topic of the contribution from the years 2021 - 2023.
A chapter on discusion of results is missing.
The discussion should also focus mainly on a critical evaluation of one's own research results.
In conclusion, there is a lack of critical assessment of the results achieved.
Author Response
Thanks for your comments!
New references are cited in this paper, and the critical assessment of the results achieved and future work are illustrated in the paper's Conclusions.
Reviewer 3 Report
-The similarity rate remains high (>20%). The text needs to be refined and references cited.
-The algorithm presented in the text should be placed in an appendix.
-Legends for certain figures should be more explicit
-Figure 16 could be replaced immediately following figure 14.
-The similarity rate remains high (>20%). The text needs to be refined and references cited.
-The algorithm presented in the text should be placed in an appendix.
-Legends for certain figures should be more explicit
-Figure 16 could be replaced immediately following figure 14.
Author Response
Thanks for your comments!
New references are cited in this paper and the algorithm presented in the text is placed in the appendix.
Figure 14, Figure 15 and Figure 16 show the energy consumption of DDPG, PPO and PPO with Reward Scaling respectively.
Reviewer 4 Report
The research presented in this paper is based on a single-train model, and this point should be clearly stated in the Introduction of the paper. While the single-train model approach forms the basis of a multi-train model, its applicability to real-life situations as the constraint of headway (a very important metro operating parameter) can only be truly related through a multi-train model.
It would be good if a short paragraph in the paper's Conclusions could be included to describe how the results can be used to support multi-train model applications.
Author Response
Thanks for your comments.
Our future work will be the multi-train operation control model. The method in this paper can provide the primary control method selection for the multi-train tracking model described in the paper's Conclusions.
Reviewer 5 Report
The article is recommended for publication. There are no significant comments. The article presents high-level scientific results that are well designed and presented.
The level of English is at a high level. There are no comments.
Author Response
Thanks for your comments!
