A Multi-Mode Recognition Method for Broadband Oscillation Based on Compressed Sensing and EEMD
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe paper presents a method for the modal analysis of broadband oscillations which occur in the power system due to the application of wind power generation equipment and power electronic devices. The signal data is collected by the Phasor Measuring Unit devices and compressed in a sub-station using the Compressed Sensing (CS) technology. The compression is recommended due to a limited communication bandwidth between the substations and the main station. Then in the main station the signal is decompressed using the Subspace Pursuit (SP) algorithm. Finally, the signal is analyzed using the Ensemble Empirical Mode Decomposition (EEMD) algorithm.
The paper is interesting and the presented method may be useful in analyzing the power system performance.
General comments:
1. How long does it take to compress and reconstruct the recorded signal? Will signals from multiple substations be able to be continuously recorded and transmitted to the main station without significant delays?
2. In Section 4.2 the results of the analysis of the signal with a very high sampling rate are presented. Can most of the WAMS devices installed in the power system record the signal with a satisfactorily high sampling rate, or is it necessary to install new devices?
Specific comments:
1. Abbreviations should not be used in the title of a paper unless necessary.
2. On page 3, the end of the first sentence reads:
"and finally transmits the frequency and amplitude about signal to main-station".
Maybe it would be better to write:
"and finally transmits the information about the signal frequency and amplitude to the main station".
3. At the end of page 5, at the algorithm point (2) there is a small letter phi, the meaning of which is not explained in the text.
4. At the end of page 5, at the algorithm point (3) the method of calculating the value of Theta with an asterisk and the value of ri-1 is not explained.
5. On page 6, at the algorithm point (6), the notation xik suggests exponentiation. At the end of this point, xi appears to be subscripted but should probably be multiplied by Theta.
6. On page 6, in the algorithm point (7) there are unnecessary letters "ddd".
7. At the end of page 6 there is "6t" and at the beginning of page 7 there is "h". Instead, it should probably be "i-th".
8. On page 7, line 1, there should be rm(t) instead of ri(t).
9. On page 9, Fig. 4 there is an error. For the signal with SNR = 0, the RE error cannot be equal to 0, because a zero SNR value means that the signal amplitude is negligibly small in relation to the noise amplitude. For a noiseless signal, the SNR value should approach infinity. The "Noiseless" point should be placed on the graph to the right of the X-axis.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors preset an interesting paper about the improvement of broadband oscilation signal, for which they propose new approaches, promising interesting results. However, it is recommended that you make some improvements to the article, namely:
1. Structure of document: Although the document follows a logical structure, the transitions between sections could be more “logical and interconnected” (greater fluidity of text and themes), especially between methodology and results.
2. Atualidade do subject: The topic is current and relevant due to the increasing integration of renewable energy sources and modern energy networks/grid advances. Broadband oscillation and its detection remain critical issues to ensure the stability of power systems.
However, the authors could have better highlighted the novelty and contribution of their work. It would also be interesting for the article to include comparisons with works that use AI or ML to detect this type of oscillation.
3. Contribution: The proposed combined CS-SP and EEMD method is innovative and effectively addresses the challenges of data compression, reconstruction, and flicker detection.
However, practical implementation challenges, such as real-time and computational requirements, must be better discussed.
4. Improves: Authors should pay particular attention to improving the following point discussions:
- Computational complexity of the proposed methods (CS-SP and EEMD) in real-world applications.
- Clarify how the method quantitatively compares/behaves with existing techniques rather than stating its effectiveness.
- Add more intuitive explanations or visuals so that readers better understand the implications of broadband fluctuation.
5. Discussion: Generally the discussion on the results, including noise robustness and signal reconstruction accuracy, is detailed and well presented.
However, the discussion on the implications of the conclusions in practical scenarios is limited and not very in-depth. For example, what is the potential for integration with other (real) modern monitoring systems?
6. Bibliography: Some references seem outdated (about half are over 10 years old), with only a few from the last two years. The bibliography could also include more recent works on AI-based solutions and advanced data analysis methodologies in energy systems.
Finally, congratulations on your work!
Comments on the Quality of English LanguageThe writing is generally clear and well-structured in terms of concepts. Some sentences, especially in the introduction and methodology sections, are very complex, so the reader would have everything to gain if the authors simplified them.
The document also has small grammatical typos.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors generally met the recommendations made previously, producing significant improvements to the article, so it is my understanding that it has the necessary quality to be published.