Machine Learning Estimation of Battery Efficiency and Related Key Performance Indicators in Smart Energy Systems

Round 1

Reviewer 1 Report

This work instead proposes a Machine Learning (ML) estimation of battery performance indicators only from time series input data. For this purpose, a random forest regressor has been trained using real data of electricity grid frequency evolution, household power demand, and photovoltaic power generation. The article has an interesting approach and presents relevant results, however there are some points that should be improved.

1- The authors mention smart energy systems in the title of the article, in the abstract and in the keywords, however in the text this is mentioned twice in the introduction, but without explaining what this refers to. I suggest checking the references that deal with Smart Energy, Internet of Energy, Energy Cloud, which are some of the terms that refer similarly to this concept, and insert a paragraph in the introduction that explains this and how the research in this article fits into this context: https://doi.org/10.1109/TSTE.2013.2288937 ,  https://doi.org/10.1002/er.8094 , https://doi.org/10.3390/su12229686

2- The authors did not clearly describe the contributions of this research in the introduction.

3- Why did the authors use Machine Learning over other techniques? This needs to be explained in the article.

4- In section 2.4 the authors bring the KPIs of the survey. Like comment one, it is necessary to conceptualize KPI and show the importance of researching them in this article.

5- What were the research limitations? Authors should make this clear in the article.

Author Response

Please, find responses in the attached file

Author Response File: Author Response.docx

Reviewer 2 Report

The article deals with the development of a novel method of various battery performance indicators using machine learning. The authors consider an important problem, hence it seems be interesting and useful for many researchers.  The article is extensive and its structure is correct. The abstract contains all the necessary information presented in an essential form. The introduction is of an appropriate length and references sufficient number of sources, most of them from the last few years, so the originality and novelty of the research has been appropriately exposed. Further on, the authors presented the background of their considerations, assumptions for the research and all necessary information on the methods applied. Later in the manuscript, the authors presented the results of the research. The presentation of the results does not raise any concerns. The authors correctly analyzed the results. The approach and the results presented in the study look credible. The work ends with accurate conclusions of significant value. The authors very aptly noted that "ML techniques are a very powerful tool at hand but does not replace an in-depth analysis and domain knowledge based discussion of results altogether".  The paper has been prepared carefully.

A detailed remark:

Formulas and descriptions in the figure 9 are too small and thus illegible. Use larger fonts.

Author Response

Please, find responses in the attached file

Author Response File: Author Response.docx

Reviewer 3 Report

Let me start by saying that the article is very interesting and, in my opinion, valuable. It touches on the problem of operation and operation of energy batteries in the smart grid. The topic is topical as can be seen from the other publications cited in the text - most of them are from the last few years. Besides, it is to be expected that the subject of energy storage will only gain in value.

The introduction to the article is correct. The abstract is good. 

The introduction should emphasize more the contribution of the article and indicate more clearly the purpose of the article.

In the remainder, the chapters are logical, mostly written clearly, although they require the reader to have quite advanced knowledge. 

The results of the experiments, presumably simulations, are presented clearly and understandably.

The discussion chapter, which is quite long and " substantive," is good. 

Literature, as previously mentioned, selected correctly.

I believe that with minor corrections in the introduction, the article can be published.

Author Response

Please, find responses in the attached file

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors satisfactorily met the recommendations of this reviewer. The article can be accepted for publication.