A Data-Driven Path-Tracking Model Based on Visual Perception Behavior Analysis and ANFIS Method
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe research aims to develop a data-driven human-like driver model for vehicle path tracking that can reflect human driving skills and behaviors, such as compensation control, preview behavior, and anticipation ability.
The topic seems original and highly relevant, especially in the context of autonomous vehicle development. It addresses the gap of incorporating human-like driving behaviors into path tracking models, which is crucial for the coexistence of autonomous vehicles and human-driven vehicles.
The summary does not provide enough detail to suggest specific improvements or additional controls for the methodology. So they need to be more detailed or informative in abstract.
A comparison by previous methods should be done. such as combination of this control with artificial intelligence.
An artificial intelligence based data-driven approach for design ideation
Development of a New Control System for a Rehabilitation Robot Using Electrical Impedance Tomography and Artificial Intelligence
also with some analyses:
Data-driven trajectory-based analysis and optimization of airport surface movement
Static analysis of a 3-RRS and a 3-RSR Spherical Parallel Robots
A modified ALOS method of path tracking for AUVs with reinforcement learning accelerated by dynamic data-driven AUV model
A new underactuated mechanism of hand tendon injury rehabilitation
some minor corrections are needed: as an example, 8lines caption of figure 2 is not good. you can add that as explanation in text.
The conclusions appear to be consistent with the evidence and arguments presented, affirming the effectiveness of the proposed human-like driver model in path tracking.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Comments and Suggestions for AuthorsThis paper is written well: Good introduction, proper methodology and experiment design, and easy-to-understand presentation.
Those are small suggestions which authors can think about
1. On page 2, lines 52-77, authors classify the modeling methods three types: classical control, modern control, and intelligent control. However, these refer to control methods rather than modeling methods.
2. The driving data for training and testing the proposed system were collected for one person, 26-year-old female. That is, the proposed system has been trained for this woman only. How about collecting data for several people and analyzing, if any, the performance difference between them. (Or authors can think that data for only one person is enough)
3. When collecting data, it seems that the human driver controls the steering angle under the given several velocities, 20km/h, 30km/h, and so on. But, in the real situation, a driver controls the pedal/brake as well as a steering wheel. If the only purpose of this research is to mimic the human's behavior under the specific situation (like driving on the curved roads), it is okay. However, as mentioned in the Introduction part, if the ultimate goal of this kind of research is to utilize the developed system in the mixed environment of human drivers and AI drivers, the authors can take both the steering angle and pedal/brake acton into account. I think it can be the future of this work.
4. This comment is a kind of extension of comment 3. Generally, Conclusion part contains the discussion on the limit of the current research and proposes the future works, which I think is missing.
5. I am not sure, but the figures and tables should be presented after the first related statements. If any figure appears but the reader has not met any statement on the figure while reading the paper, he/she may be perplexed.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
Comments and Suggestions for AuthorsIn this study the authors present a data-driven path tracking model. Particularly, the authors have focused on the use of visual perception behavior analysis and the ANFIS method. The topic is interesting and the specifications of the study are well presented. Below, I quote some suggestions that I believe could improve the overall quality of this manuscript:
In the abstract, the authors should better highlight the research gap that their study aims to solve. Additionally, they should present their contribution and some of their main findings.
In the introduction section, the authors clearly present the background of the study and highlight the research gap and their main contribution.
Section 2 presents the human-like driver model and the steering control. The figures used are easy to read and understand and certainly add value. The layers and specifications of the suggested model are clearly described. Nonetheless, I believe that the authors should include more information about other studies that have used similar approaches and methods.
In Section 3, the authors clearly define and present the human driving data collection process that they used. Additionally, they have adequately provided the required specifications regarding their procedures.
In the result analysis section, the authors go over their results in detail. However, the outcomes should be further explained and comparisons with the outcomes of other related studies must be included.
One of the main issues of this manuscript is the lack of a constructive discussion. The authors should include a discussion section in which they go over their suggested model, further analyze their results, make connections to the existing literature, highlight the potential implications of their approach, and comment on the limitations and future extensions of their algorithm.
The conclusion section should also be further expanded. The findings should be better highlighted and explained, the implications of the model should be presented, and directions for future research should be given. It should be clearly specified why someone would select the suggested model instead of other existing ones. Finally, the authors should provide some conclusive remarks about their study, again, while taking into account other related studies.
Comments on the Quality of English LanguageThe quality of English is satisfactory. Only some minor corrections are required.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Round 2
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors have satisfactorily addressed all the previous comments and suggestions made. I believe that the overall quality of the manuscript has been improved.
Comments on the Quality of English LanguageThe quality of English is satisfactory. The authors have made improvements throughout the text.
