Round 1

Reviewer 1 Report

Can improve the abstract for the manuscript.

Cite more relevant papers/articles

Comments for author File: Comments.pdf

Author Response

The authors would like to hereby express their sincere gratitude to the respectable reviewer for their efforts in reviewing our paper, and providing us with constructive feedback. The reviewer’s concerns and raised issues are fully addressed in this response letter. We believe we have addressed all the comments raised by the respectable reviewer in the following.

In this letter, the comments of the reviewer and our responses are indicated in red and black, respectively. Also, all the changes in the revised paper are highlighted in blue.

1- Can improve the abstract for the manuscript.

The abstract is updated in the revised version and is highlighted in blue in the revised manuscript.

2- Cite more relevant papers/articles

New related references added in the revised version are: [3], [5], [17], [18], [25], [26], [32]

Reviewer 2 Report

The paper discusses an interesting metaheuristic, based estimation technique in the field of battery management for EV sector. This is a rapidly emerging complex space. It is important to highlight the key merits of any algorithm proposed in this space with respect to how easy it is to adopt, implement and safety aspects. Mission critical algorithm design is usually based on robust traditional controls running on edge devices for various reasons. Any application which correlates to the safety of an EV shall need to justify the potential of the algorithm highlighting and addressing the key parameters which relates to accuracy, safety, throughput, failure rates and maintainability. While the paper is well formulated and structured with relevant design facts and results, I would recommend establishing the merits of this proposal highlighting the pointers as mentioned above. This will help the paper stand out and differentiate from other available material in literature. Table 1 is a very good summary and documents most of the details and can be used to add a few more details as requested in the above section.

 

1.     Indicate all current methods that are used for the robust estimation and address the positives and negatives of those methods. Then compare what are benefits of using the proposal in this paper.

2.     A few more relevant references must be surveyed and reported as needed. This space is huge and shall be studied appropriately to cite all relevant materials online. A few are listed below:

·      Theory of battery ageing in a lithium-ion battery: Capacity fade,

             nonlinear ageing and lifetime prediction

-        https://doi.org/10.1016/j.jpowsour.2020.229026

·      Understanding ageing in Li-ion batteries: a chemical issue

             - M. Rosa Palacin

·      A comprehensively optimized lithium-ion battery state-of-health estimator based on Local Coulomb Counting Curve

-       https://doi.org/10.1016/j.apenergy.2022.119469

·      A review on online state of charge and state of health estimation for lithium-ion batteries in electric vehicles

-       https://doi.org/10.1016/j.egyr.2021.08.113

·      Online Estimation of Lithium-Ion Battery Capacity Using Deep Convolutional Neural Networks

-       https://doi.org/10.1115/DETC2018-86347

 

3.     Please expand elaborate SoF, SoP in the introduction section as it appears for the first time in the description.

4.     In Figure 1, please include more details on the battery back, number of parallel/series configuration with voltage and power ratings. Please include all output interfaces (temperature, current, voltage, Magnetic field) which are available to BMS for processing.

5.     What is the nominal operation of the DFN model? Does it use a CC-CV charge/discharge cycle? Does the internal temperature of the battery affect the charge/discharge cycle? It may not be relevant to the study in this proposal but a paragraph stating the nominal controller design will help here.

6.     Please cite references to Table 3. I believe it is taken from [31,32], if yes please add that in the place where Table 3 is referenced.

7.     Nit! – There is an unlinked citation for Equation 6 after line number 234. “is pooling area size [? ].”

8.     Figure 3. Is a very good illustration of the CNN architecture. What is the non-linear activation after each conv layers and what is the output activation function? What is the error loss function used and the optimizer?

9.     Description of the CNN layer is inadequate. Please explain the numerical details of the conv filter dimensions, number of filters, stride, etc. That would help understand the model a bit better. My calculation says filter size is 3x3. If that is true, please document and why do you think that is justifiable. What is the input layer structure – [30,22,1]?

10.  What about discharge rates? Does it have a factor to play in the process as well? Why is data taken from charge cycles only? Also, what about very low C rate charging?

11.  A tabulated representation, of the CNN results would be nice to have for the offline training results. What is the training time, accuracy, how the loss evolved, best epochs, etc.

12. Conclusion section can be improved, with a clear statement of what is new, how it is beneficial and what can be improved in future.

 

 

 

 

 

 

Author Response

The authors would like to hereby express their sincere gratitude to the respectable reviewer for their efforts in reviewing our paper, and providing us with insightful, thorough, and constructive feedback. The reviewer’s concerns and raised issues are fully addressed in this response letter. We believe we have addressed all the comments raised by the respectable reviewer. In doing so, we have crafted an article (attached as a PDF) that is more rigorous in the content and clearer in the presentation.

In this letter, the comments of the reviewer and our responses are indicated in red and black, respectively. Also, all the changes in the revised paper are highlighted in blue. Please find the attachment.

Thank you.

Author Response File: Author Response.pdf