Prediction of Battery Remaining Useful Life Using Machine Learning Algorithms
Round 1
Reviewer 1 Report
This research article discusses the prediction of battery remaining useful life (RUL) using machine learning algorithms. It categorizes the methods into model-dependent and data-driven techniques and compares the performance of different machine learning algorithms. The results show that the random forest algorithm performs the best in terms of prediction accuracy.
1.What are the advantages and limitations of model-dependent and data-driven techniques for predicting battery remaining useful life?
2.What are some potential areas for future research in improving the prediction accuracy of battery remaining useful life using machine learning algorithms?
3.The present study has several limitations in its research methodology.
Firstly, the study relies on historical data for training the machine learning models. This can introduce biases and limitations in the accuracy of the predictions, as the quality and representativeness of the historical data may vary.
Secondly, the selection and optimization of hyperparameters for the machine learning models require human judgment. This introduces subjectivity and potential biases in the model training process.
Thirdly, the study does not explicitly address the issue of imbalanced datasets, where the number of samples from different remaining useful life (RUL) classes is uneven. This can affect the accuracy of RUL prediction, as the models may be biased towards the majority class.
Lastly, the study does not consider the uncertainty associated with RUL predictions. Estimating and quantifying the uncertainty can provide more reliable and informative results for decision-making.
How to resolve these issues?
4. The first and second paragraphs of the introduction require a large amount of references support, and additional references should be added.
For example, “Since majority of end user electronics are motorized by battery-like and the use of renewable energy sources to generate electricity is expanding quickly, energy storage has emerged as one of the key sectors. Due to its high energy density and lengthy cyclical and calendrical lifetime, lithium-ion batteries (LIBs) as efficient energy storage systems have assumed leading position in powering EVs.” The following references are recommended.
Journal of Energy Chemistry 2023, 80, 625-657; InfoMat. 2022, 4(11):e12365; Small Methods 2018, 2(11): 1800156; J. Am. Chem. Soc. 2022, 144(32), 14638-14646; Adv. Funct. Mater. 2022, 32, 2205471
Comments for author File: 
Comments.pdf
Author Response
We are enormously grateful for your time and the valuable feedback we have received. We have tried to do our best to respond to questions and suggestions, incorporating the necessary modifications to our revised article. Without a doubt, thanks to your comments, we have been able to improve this version and we are sure that it meets the high-quality standards of your prestigious journal.
In yellow we have highlighted the modifications that have been incorporated into the revised version of our article and commented on in this response document to the reviewers.
Sincerely,
Ernesto DR Santibanez Gonzalez, PhD
Corresponding Author
Author Response File: Author Response.pdf
Reviewer 2 Report
This paper proposes a novel remaining useful life (RUL) prediction framework using several machine learning methods. The following concerns need to be addressed:
1. The term "version-based" mentioned on page 3 needs to be clarified.
2. Further proofreading is required. Please correct all typos and grammatic errors.
3. It is suggested not to start a paragraph with a reference index, e.g., on page 2.
4. The figure quality needs to be improved for better readability.
5. Please provide detailed explanation of your prediction method and highlight the novelty of your RUL prediction framework in the Introduction.
6. Table 6 needs be rearranged.
7. It is suggested to include a comparison of your method with other methods mentioned in the introduction to demonstrate the pros and cons.
Needs to be improved.
Author Response
We are enormously grateful for your time and the valuable feedback we have received. We have tried to do our best to respond to questions and suggestions, incorporating the necessary modifications to our revised article. Without a doubt, thanks to your comments, we have been able to improve this version and we are sure that it meets the high-quality standards of your prestigious journal.
In yellow we have highlighted the modifications that have been incorporated into the revised version of our article and commented on in this response document to the reviewers.
Sincerely,
Ernesto DR Santibanez Gonzalez, PhD
Corresponding Author
Author Response File: Author Response.pdf
Reviewer 3 Report
This paper presents a prediction model for battery RUL using machine learning algorithms. The authors highlight the importance of accurately predicting RUL for effective battery management and provide a comprehensive review of relevant literature. They propose a framework that involves data extraction and preprocessing, model training, and prediction and investigation. The proposed methodology is evaluated using a real-life battery dataset and performance evaluation metrics.
1. In Section 1, there are discrepancies in the reference citation format. Please ensure that all references are cited consistently and according to the journal's prescribed guidelines.
2. I recommend a thorough check of the entire manuscript for formatting and spelling errors. For instance, the formatting issue in Table 6 should not have been present.
3. The title of the manuscript appears too broad.
4. Only the various Machine Learning algorithms models are used in this paper. The presented results seem trivial and the contribution is therefore questionable. Compared with recent Machine Learning based remaining useful life prediction approach such as Energy, 10.1016/j.energy.2023.128442; Measurement Science and Technology, DOI: 10.1088/1361-6501/ace072; Quality and Reliability Engineering International, DOI: 10.1002/qre.3314, authors need to do a better job of highlighting their contributions.
5. With regards to the remaining useful life of batteries, some references used in this manuscript appear to be outdated. I recommend incorporating more recent citations, preferably from the past three years, from this journal or other relevant journals.
6. The clarity of the figures and equations throughout the manuscript needs enhancement. Please ensure that all images and formulas are of high resolution and easily legible.
7. Can you provide more insights into the potential applications and implications of your research findings?
Author Response
We are enormously grateful for your time and the valuable feedback we have received. We have tried to do our best to respond to questions and suggestions, incorporating the necessary modifications to our revised article. Without a doubt, thanks to your comments, we have been able to improve this version and we are sure that it meets the high-quality standards of your prestigious journal.
In yellow we have highlighted the modifications that have been incorporated into the revised version of our article and commented on in this response document to the reviewers.
Sincerely,
Ernesto DR Santibanez Gonzalez, PhD
Corresponding Author
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The authors have generally addressed all major concerns in the last round of review. I have no more comments.
Author Response
Enclosed
Reviewer 3 Report
After reviewing the revised manuscript, I find the corrections satisfactory and recommend its acceptance for publication.
Author Response
Enclosed
Author Response File: Author Response.pdf
