Deep Reinforcement Learning Algorithm Based on Fusion Optimization for Fuel Cell Gas Supply System Control

Round 1

Reviewer 1 Report

In this work, authors discussed that the Deep Reinforcement Learning Algorithm Based on Fusion Optimization for Fuel Cell Gas Supply System Control. The work is interesting. However, the following comments needs to be addressed.

11.       Authors mentioned that the undecoupled 7 FO-DDPG algorithm has faster dynamic response and more stable static performance compared to the fuzzy PID, DQN, traditional DRL algorithm. Is there any evidence for the statement?

22.       In the introduction section, I suggest the authors give a brief introduction to fuel cells prior to PEMFC.

33.       The conclusion part is wages. Please discuss some more about PEMFC.

44.       There are some relevant recent papers that can be included in the References section: doi.org/10.1016/j.energy.2021.121791.

Author Response

Response to Reviewer 1 Comments

Dear Reviewer:

We thank the reviewers for your careful review of the paper and insightful suggestions. We have made every effort to revise and improve the paper to address the questions. The responses (in red) and details (in blue) of revisions for each issue are explained below.

We hope this revised manuscript has addressed your concerns, and look forward to hearing from you.

Sincerely yours,

Hongyan Yuan, Zhendong Sun, Yujie Wang, Zonghai Chen

Corresponding Author:

Professor Zonghai Chen

E-mail: [email protected];

Address: Department of Automation, University of Science and Technology of China, Hefei, 230026, China.

Tel: +86 055163606104

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper studies a control algorithm based on deep reinforcement learning to effectively control the air and hydrogen supplies in a PEM fuel cell system. The discussions and results should attract audience interested in fuel cell electric vehicles. Here are some suggestions to improve the paper.

1. Reference number on line 96 is missing. Line 181, the direct quotation should have a reference.

2. It is unclear how equation (1) is obtained. More details will be needed. I think E is the open circuit voltage, and the authors may want to use these terms in the field of fuel cells.

3. For all the parameters, units should be specified.

4. For equations (2) and (3), how are the overpotentials obtained in this study? It is unclear.

5. On line 107, the second point of contribution is unclear. Where does this reflect in the manuscript?

6. Below line 231, what does ‘economic efficiency’ mean? This needs to be defined.

7. One line 235, why this is called peroxide oxygen ratio? This term is also used in the manuscript later.

8. It is unclear how equation (13) was obtained.

9. Figure 3, descriptions should be added for each sub-figure on the caption.

10. On table 3, we can see that the best algorithms for HER and OER are different. Why is that? Does it depend on the concentrations of the gases, or other parameters? This needs to be discussed thoroughly in the manuscript.

Author Response

Response to Reviewer 2 Comments

Dear Reviewer:

We thank the reviewers for their careful review of the paper and insightful suggestions. For each of the points made, we have appropriately modified the manuscript where possible. The responses (in red) and details (in blue) of revisions for each issue are explained below.

We hope this revised manuscript has addressed your concerns, and look forward to hearing from you.

Sincerely yours,

Hongyan Yuan, Zhendong Sun, Yujie Wang, Zonghai Chen

Corresponding Author:

Professor Zonghai Chen

E-mail: [email protected];

Address: Department of Automation, University of Science and Technology of China, Hefei, 230026, China.

Tel: +86 055163606104

Author Response File: Author Response.pdf