Fusion Technology-Based CNN-LSTM-ASAN for RUL Estimation of Lithium-Ion Batteries
, Xiaojuan Qin 1, Furong Ma 2, Haoran Wu 1, Min Chai 1, Fujing Zhang 1, Fenghe Jiang 1 and Xu Lei 1Round 1
Reviewer 1 Report
Comments and Suggestions for Authorssee the attachment
Comments for author File: 
Comments.pdf
good
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors1. The article mentioned the convolutional neural network (CNN) did not discuss the applicability of the method of the method with the advantages and disadvantages, but the other methods are discussed, it is recommended to add
2. The size and characteristics of the dataset for experimental comparison are not described by the authors, is it possible to add this part of the content as a supplement?
3. The CNN-LSTM-ASAN model proposed by the authors is compared with CNN-LSTM, LSTM-ASAN and other models, and other models are not described here, so it is suggested that the authors add references.
4. It is recommended that the authors add a comprehensive comparison with other advanced models, including but not limited to other deep learning models, when comparing models.
5. It is recommended that the authors provide more metrics for comparison, not only RMSE and MAPE, including but not limited to precision, recall, F1 score, and so on.
6. The authors consider the effects of different environmental factors on the model, such as temperature, charging and discharging, and suggest that the authors add a discussion of how these factors affect the predictive performance of the model.
7. For the design of model parameters, it is suggested that the authors provide more details on the discussion of parameter settings.
8. Some of the graphs in the article are small and unclear in terms of horizontal and vertical coordinates, figure legends, text, etc. Please check and correct the authors.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for Authors In the manuscript titled ‘Fusion Technology-based CNN-LSTM-ASAN for RUL Estimation of Lithium-ion Batteries‘, the authors present an AI approach to predict the capacity decay of the batteries. Though the study's objectives are not entirely novel, the study attempts to improve RUL prediction of batteries by introducing a combinatorial approach of CNN and LSTM with ASAN mechanism. The results are promising and the manuscript can be considered for publication if the following information is clarified:1. Description and labels within Figure 1 are not readable. Since CNN and LSTM are described in separate images, their schemes in figures 1 can be summarized.
2. It would be insightful if little information about the batteries is shared in the paper. Are these intercalation batteries? Can the present results be linked to battery chemistries? What are the electrochemical phenomena where this model is likely to fail?
3. In section 3, (On Page 7, line 188) authors describe datasets by a number (#B05, #B07, #B32 etc). Can authors provide a table with the necessary details of these datasets and their sources? It would be easier to compare.
4. Are the models used for comparison in section 4.2 from previous studies? They have not been cited.
5. What are the differences in the 5 models compared in Table 4. Knowing the technical differences will highlight strengths of present approach.
6. To allow reproducibility and reuse (FAIR) of the technology, are authors planning to share the associated codes (models, data pre-processing, etc) open source?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
