Clustering Analysis of Voltage Sag Events Based on Waveform Matching
Round 1
Reviewer 1 Report
The authors of this article present the results of the analyzes regarding voltage sag detection and classification. The novel method for voltage sag proposed by the Authors was tested on data from 10 kV points in Beijing, China. The article emphasizes the need to perform such analyzes from a technical point of view, which is to improve the operation of the network. However, there are some considerations and questions that should be addressed and clarified:
• The literature review should be more extensive and in-depth. The list of literature provided in the article is too poor.
• The authors should provide more information about the 10kV grid.
• At what points in the network were the data collected for analysis?
• Why was local sudden-drop point proposed?
• In my opinion, formulas should be numbered, eg page 4 - formula X =…. - no numbering; page 10 - formula wij = ... - no numbering, page 11 - formula L = ... - no numbering, page 12 - formula Urms (k) = ... - no numbering,
• At the end of some lines there are single letters "a" or the articles "the". They should appear at the beginning of the next line.
• Figure 1, Figure 2, Figure 3, Figure 4 - units are missing from the graph axes.
• Figure 4, Figure 7, Figure 11, Figure 12, Figure 13 - too small font in the diagrams.
• Page 10 - in the presented algorithm, individual lines of text are too close to each other (too small space between lines).
• Units are missing from the tables.
• Conclusions should be more elaborate. In particular, it should be described what new the presented article brings in comparison with other works of this type. In the current version, the summary describes what was done in the article.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
How do we analysis the data results listed in Tables 2 to 5 ?
The occurrence of sag can be known by RMS calculation as well, what is the advantage of detecting sags with wide detecting range by the proposed method?
The current data dealt with only voltage sag. What do you think the performance of the proposed algorithm for voltage swell?
It seems the title is not specific. Please change the title of paper to clearly say the main idea of the proposed scheme.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Thanks to the authors for their responses. I have no more questions.
Author Response
Response to Reviewer 1 Comments
We would like to thank the editors and the reviewers for their advice and comments regarding our manuscript entitled " A New Method for Voltage Sag Detection and Classification" (processes-1585134).
Reviewer 2 Report
The authors do not suggest specific reason or explanation for my answers.
Please reanswer my questions.
Author Response
Response to Reviewer 2 Comments
We would like to thank the editors and the reviewers for their advice and comments regarding our manuscript entitled " A New Method for Voltage Sag Detection and Classification" (processes-1585134). All these advice and comments are really valuable and helpful in revising and improving our paper, as well as an important guide significant to our research. We have studied them carefully and have made our changes accordingly. The revised parts according to the comments are marked in red. The main changes in the manuscript and our responses to the comments are as follows.
Point 1: How do we analysis the data results listed in Tables 2 to 5 ?
Response 1: Thanks for your comments. Table 2 shows the coordinates of the start point to the end point of the voltage sag detected using different methods. We can conclude that the sag segments detected by RMS are all inside the abnormal segment detected by FPAS, therefore FPAS is more sensitive than method RMS.
Table 3 shows the cluster analysis results of the A-phase voltage data.Table 3 is an affinity matrix calculated using the abnormal waveform segment of the A-phasevoltage data. The closer the value in Table 3 is to 1, the higher the similarity between the two curves. Tables 4 and 5 show similar results.
Point 2: The occurrence of sag can be known by RMS calculation as well, what is the advantage of detecting sags with wide detecting range by the proposed method?
Response 2: Thanks for your comments. We present a new method for extracting the voltage anomaly waveform, which is a fast projection segmentation algorithm (FPSA). From Table 2, we can see that the sag segments detected by RMS are all inside the abnormal segment detected by FPAS, therefore FPAS is more sensitive than method RMS. The abnormal voltage waveform detected by the FPAS method contains more information, so it is natural to have a better performance in the cluster analysis.
Point 3: The current data dealt with only voltage sag. What do you think the performance of the proposed algorithm for voltage swell?
Response 3: Thanks for your comments. Currently we only collect data on voltage sags, so only the voltage sags are analyzed. From a data analysis perspective, data transformation can be used to convert the voltage swell waveform into a voltage sag waveform. Therefore, the method proposed in this paper should also perform well on voltage swells.
Point 4: It seems the title is not specific. Please change the title of paper to clearly say the main idea of the proposed scheme.
Response 4: Thanks for your comments. According to your suggestions, we use the new the title of paper as ‘Clustering Analysis of Voltage Sag Events Based on Waveform Matching’.( See title page of revised manuscript)
We have done our best to improve the presentation of the manuscript. We have marked all our changes in the revised manuscript.
Once again, thank you all very much for your comments and suggestions which led to our further study on the topic and our opportunity to improve the manuscript.
Round 3
Reviewer 2 Report
-
Author Response
Please see the attachment
Author Response File: Author Response.pdf
