Game-Based Flexible Merging Decision Method for Mixed Traffic of Connected Autonomous Vehicles and Manual Driving Vehicles on Urban Freeways

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors, you have chosen a very relevant topic for research. I have the following recommendations and questions:

1. "flexible" is written in the Title of the manuscript. The authors do not use this term in the Absract, Conclusions, etc. I suggest to agree

2. I recommend noting "the study acknowledges several limitations" in the Abstract

3. "3.1. Incomplete Information Static Game Models" is the development of the authors?

4. How to understand "Incomplete Information"?

5. Has the simulation been tested? Where are the results?

6. Since not everything is clear on the graphs, I recommend presenting the main results in a table. For auxiliary results in the form of a table, you can make an Appendix.

Kind regards

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Title: The title of the article corresponds to the topic of the article.

Abstract: The abstract is devoted to describing the methods used to conduct the experiment. In this regard, it is suggested that the authors define the objective of the conducted study in the abstract, then the study area for conducting the experiment, and then define the obtained results more precisely. This will make the abstract comprehensive and clear.

Introduction: The introductory part of the work is too concise and incomplete. It seems that the authors focused more on decision-making methods and driving styles, without highlighting previous studies that dealt with the same research. Namely, it would be important to highlight, in addition to the description of the methods, the results of the studies that applied them in their research, pointing out their advantages and disadvantages.

Problem Description and Framework: The authors have clearly indicated the considered problem and the architecture of the research.

Flexible Merging Decision Methods: For the purposes of this research, the authors applied different methods and strategies, but they paid more attention to some and less to others. Namely, the suggestion is that the authors emphasize the essence when explaining the methods concerning the research itself. For example, it would be good if the DRL-based Launch Decision chapter described the methods more concisely or perhaps presented them in a table, leaving room to describe e.g. Nash equilibrium strategies.

Dynamic Driving Style Recognition Method: To begin with, bearing in mind the problem under consideration, it would be useful for the authors to highlight the flow of vehicles on the considered section of Miyun Road of the Tianjin City Expressway. This would further underline the essence of the problem and put the focus on possible conflict situations due to different traffic conditions and driving styles of drivers. Also, what vehicle categories were considered? Next, the authors should provide a more extensive explanation on the question of why the distribution of data for training and testing the SVM model. Also, it would be good to highlight again the data used for training and its sample.

Simulation Experiments: Which neural network was used to train the DRL model? What do 128 neurons represent in the input layer of a neural network, and what do 13 neurons represent in the output layer? Also, in addition to displaying graphs (Figures 6 and 7), the authors should provide a more precise description of the obtained results. This is because it is not clear to what extent the rewards obtained by CAVs through learning training when combining different driving styles are significantly enhanced and stabilized. Also, looking at Figures 9 and 10, I notice that there are multiple graphs in one image that are not adequately labeled, so in that regard, it is much more difficult to follow your explanation of the results without specifying which graph it refers to. The suggestion is to number the graphs e.g. (a), (b), etc. and you highlight it in the description of the results when you give an explanation.

Conclusions and Future works: Before reaching the conclusion, it is suggested that the authors discuss a little about the results of earlier research (on the same topic) comparing them with their own results. In this way, it is possible to highlight either the method used or the way to conduct the research, its advantages, disadvantages, etc. Also, as with the abstract, the authors need to highlight the most important results of the research.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The article is devoted to a significant theme of the decision-making process for managing mixed traffic with Manual Driving Vehicles (MDVs) and Connected Autonomous Vehicles (CAVs) based on game theory. The advance of the article is a clear description of the problem and presents a mathematical model. Also, a lot of plots show the result of the simulation that's a strong part of the article.

There are such notes

1. In formula (3) it would be better to use different letters for the upper limit of the integral and the variable for integration.  The authors use the same variable t. It has to be changed.

2. In formula (10) the authors use function f(*), described in line 269 as a normalization process. It would be better to show it in detail.

3. In the description of formula (14) in lines 319-322, some variables are not presented in the formula (14).

4. In the description of formulas (15)-(17) there's a variable \omega (see line 342) omitted in formulas.

5. In formula (16) there are variables r and \gamma omitted in the description.

After correction, the article can be published

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors, thank you for your work

Author Response

Thank you very much for taking the time to review this manuscript again.

Reviewer 2 Report

Comments and Suggestions for Authors

-

Author Response

Thank you very much for taking the time to review this manuscript again.