PONE-D-20-19190

The Data Dimensionality Reduction and Bad Data Detection in the Process of Smart Grid Reconstruction through Machine Learning

PLOS ONE

Dear Dr. Gou,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Aug 20 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Zhihan Lv, Ph.D.

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information

 

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper introduces the method of machine learning virtual data injection attack detection, builds the attack data set through simulation, and uses MatPower tools to simulate and analyze the dataset. Using an isolated forest anomaly score data processing algorithm and a local linear embedded data dimensionality reduction method, a data feature extraction algorithm is constructed. Finally, given a combination of convolutional neural network and gated recursive network, an algorithm model for false data injection attack detection is constructed. This topic is very interesting, but the following revisions need to be made before publishing:

1: In the introduction section, the introduction to the objective and innovation of the research is very similar to the abstract. The final part of this section needs to elaborate the research significance and innovation, so the author(s) needs to reorganize this part of the content.

2: It is recommended that the author(s) add a section of the literature review after the introduction to summarize the shortcomings of the previous research on the basis of analyzing the previous research, so as to highlight the views of this paper.

3: In the method section, the basic principles of the method used in the paper need to be explained. Now, the method section in the paper is more like the experiment section. The basic principles of the isolated forest, convolutional neural network, and gated recursive network used in this paper need to be explained in the method section.

4: The current method section is suggested to be changed to the experiment section. Also, adjust the format of the letters in the text so that they are at the same height as the text.

5: When explaining the meaning of the letters in the paper, the letters should be in italic format. The author(s) should check the entire text and adjust the format.

6: Are Figures 1 to 7 missing in the text? Check the paper to ensure that the necessary chart data are retained.

7: Simulation analysis is mentioned in the results section of the paper, but the parameter settings of the method used and the operating environment are not given. Please supplement these contents.

8: There are many discussions in the paper, it is recommended to simplify them.

9: The description of the research content in the conclusion section has failed to express the method and all the research content in the paper. I hope the author(s) can adjust the summary of the research content. At the same time, the conclusion needs to include the outlook for the future to better highlight the significance of this paper.

10: It should also be noted that there should be no citations in the results and discussions of the paper because this is the author’s research results and the content of the discussion based on the research results. I hope the author(s) can think about this issue.

Reviewer #2: To detect fake data injection attacks (FDIAs) power grid reconstruction, solve the problems of high-dimensional data, and process the bad abnormal data in the power system, thereby achieving safe and stable operation of the power grid system, this paper introduces machine learning methods to explore the detection of FDIAs. First, by using the standard IEEE node system and simulated fake data to inject non-complete topology information under attack conditions, the construction of the attack data set is completed, and MatPower tools are applied to simulate and analyze the dataset. Second, based on the iForest anomaly score data processing algorithm combined with the local linear embedding (Michelle) data dimensionality reduction method, a data feature extraction algorithm is constructed. Finally, based on the combination of CNN and GRU, a fake data injection attack detection algorithm model is constructed. The results show that in the IEEE14-bus node and IEEE118-bus node systems, the estimated state distribution before and after the attack vector injection is consistent with the initial value. In the iForest algorithm, the number of iTree and the number of samples will affect the extraction of abnormal score data. When it is determined that the number of iTree n is 100 and the corresponding sample number w is 10, the detection effect of the algorithm is the best. The fake data injection attack detection algorithm model based on CNN-GRU has a good detection effect under high attack intensity, with an accuracy rate of more than 95%, and its performance is better than other traditional detection algorithms. In this paper, the bad data detection model based on deep learning has a positive effect on the safe and stable operation of the smart grid. The paper should make the following amendments before publication.

1. The author needs to use the full names instead of the abbreviations in the keyword list of the paper, which will make the paper easier to be found. Please revise this issue.

2. The introduction section of the paper describes the basic content of FDIAs in a large amount of space, but the author has not described and analyzed the current research situation worldwide in the relevant direction of the paper. The author needs to add the corresponding content.

3. The author needs to clearly point out the innovation of the research to indicate the significance of the paper. Please add the corresponding content in the last paragraph of the introduction section.

4. The author introduces the various methods that will be used in the paper in the method section, but the author does not mention how the methods are combined. Please add relevant content.

5. The author needs to add a new simulation experiment section in the method section to introduce the content associated with the experiment. At present, the author has not introduced the experiment.

6. The DBN algorithm is mentioned in Figure 7, but the author does not mention this content in the method section. Please also explain this algorithm in the method section.

7. In the discussion section of the paper, the author describes the basic content in a large amount of space, but the existence of this content in the discussion section is unreasonable. Please move the relevant content to the introduction section or delete it.

8. In the introduction, the author not only needs to briefly describe the results of the research but also has to compare the results of this research with those of other scholars and therefore draws conclusions on whether the results are consistent. If not, please explain why. Please supplement the relevant content.

9. In the conclusion section of the paper, the author needs to describe the shortcomings of the paper and the reasons for it in more detail, as well as describing the plan or prospect for future research. Please supplement the relevant content.

Reviewer #3: The authors have applied the machine learning methods to the detection of FDIAs, which is a novel point of view. The authors have completed the feature extraction of FDIAs data by constructing the FDIAs dataset in the IEEE14 bus and IEEE118 bus node systems based on the theoretical basis of the isolated forest algorithm and local linear embedding method of machine learning. In summary, the deep learning-based FDIAs data detection model explored in this paper achieves safe and stable operation of the smart grid. However, the paper needs to be further revised to meet the publication standards, as follows:

1: In the introduction section, the authors used a lot of words to introduce the harm and detection methods of FDIAs, but they did not elaborate the application and advantages of machine learning in FDIAs. So, please introduce the specific application status.

2: The explanation of the parameters inserted in the text is not displayed on the same line as the text layout, and the font size is also different. Please make revisions to these issues. In addition, the figures in the text are not displayed well, which makes readers unable to see the specific information. Please replace them with valid figures.

3: As mentioned in the paper, the GRU is combined with deep learning. However, the authors did not explain clearly the role of GRU and the design key in combination with CNN.

4: In this paper, the authors did not explain the environment and detailed parameter settings of the simulation experiment. The relevant content should be explained together with the relevant indicators in Section 2.4.

5: In Section 3.3, the authors have compared the proposed detection method with other traditional methods. A brief description of the traditional algorithms should be added in the main text to point out the shortcomings of the traditional algorithms, thereby reflecting the advantages of the proposed algorithm.

6: The repetition degree of the discussion section and the introduction section is too high. Please focus on the discussion and analysis of the results of this paper. In addition, it is not reasonable to introduce the research background in the last paragraph of the discussion. Please revise it.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: Xin Gao

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

The Data Dimensionality Reduction and Bad Data Detection in the Process of Smart Grid Reconstruction through Machine Learning

PONE-D-20-19190R1

Dear Dr. Gou,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Zhihan Lv, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I read and carefully evaluated the revised version of this manuscript. The authors have deal with the comments and suggestions of reviewers in a highly satisfactory and constructive manner. The revised manuscript clearly meets the standards that have been required by reviewers. I propose the acceptance of the manuscript.

Reviewer #2: To detect fake data injection attacks (FDIAs) power grid reconstruction, solve the problems of high-dimensional data, and process the bad abnormal data in the power system, thereby achieving safe and stable operation of the power grid system, this paper introduces machine learning methods to explore the detection of FDIAs. First, by using the standard IEEE node system and simulated fake data to inject non-complete topology information under attack conditions, the construction of the attack data set is completed, and MatPower tools are applied to simulate and analyze the dataset. Second, based on the iForest anomaly score data processing algorithm combined with the local linear embedding (Michelle) data dimensionality reduction method, a data feature extraction algorithm is constructed. Finally, based on the combination of CNN and GRU, a fake data injection attack detection algorithm model is constructed. The results show that in the IEEE14-bus node and IEEE118-bus node systems, the estimated state distribution before and after the attack vector injection is consistent with the initial value. In the iForest algorithm, the number of iTree and the number of samples will affect the extraction of abnormal score data. When it is determined that the number of iTree n is 100 and the corresponding sample number w is 10, the detection effect of the algorithm is the best. The fake data injection attack detection algorithm model based on CNN-GRU has a good detection effect under high attack intensity, with an accuracy rate of more than 95%, and its performance is better than other traditional detection algorithms. In this paper, the bad data detection model based on deep learning has a positive effect on the safe and stable operation of the smart grid. In this revision, the authors have already addressed all the comments. The article has been greatly improved, and most of my doubts have been answered in the response. There may still be some room for improvement. Anyway I think it is acceptable.

Reviewer #3: The authors have applied the machine learning methods to the detection of FDIAs, which is a novel point of view. The authors have completed the feature extraction of FDIAs data by constructing the FDIAs dataset in the IEEE14 bus and IEEE118 bus node systems based on the theoretical basis of the isolated forest algorithm and local linear embedding method of machine learning. In summary, the deep learning-based FDIAs data detection model explored in this paper achieves safe and stable operation of the smart grid. In this revision, the authors have already addressed all the comments. The paper can be accepted now.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: Xin Gao