Wavelet Analysis to Detect Ground Faults in Electrical Power Systems with Full Penetration of Converter Interface Generation

Round 1

Reviewer 1 Report

Review about: Wavelet Analysis to Detect Ground Faults in Electrical Power Systems with Full Penetration of Converter-Interface Generation.

The paper is well written, well presented, and covers an important topic. Based on this first observation, my review will be a minor revision focused on improving the quality of the presentation of the paper.

1) I miss the mathematical background of wavelets. You could include the equations to explain the wavelet transform, and how the signal is decomposed. If you are interested this could be found in equations 1 to 4 of this paper: Wavelet LSTM for Fault Forecasting in Electrical Power Grids. 

If you prefer a more deep explanation I would suggest following equations 21 to 25 of this paper: Fault Prediction Based on Leakage Current in Contaminated Insulators Using Enhanced Time Series Forecasting Models 

2) I think the presentation of the results using scientific notation would be better than the presentation used. I mean use $\times 10ˆ1$ instead of E-1 (in the tables).

3) Figures 1, 3, and 5 are bigger than the others (considering the font size). My suggestion is to resize them to a smaller size or standardize the size font of all pictures.

These are just suggestions to improve the paper, in my opinion, the research is well done. Congratulation!

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Wavelet transform analysis is proposed to detect ground faults with low current contributions in the work. Some revisions are needed.

(1) Daubechies mother wavelets dB3 and dB2 are used in wavelet transform. Why does the author use these? Why not use others?

(2) The author needs to compare the proposed method with other commonly used technique.

(3) Computation time also need to be calculated and compared.

(4) The sensor accuracy is of vital in the experiment. How do you ensure accuracy? How do you avoid noise?

(5) Why does the author not conduct actual testing?

(6) The author needs to compare the wavelet transform with other signal decomposition algorithm: VMD(10.1109/TVT.2021.3138959).

(7) Has the proposed method under extreme conditions been tested?

(8) There was too little discussion and the discussion lacked mechanism reason.

(9) Voltage threshold is set to 0.15 and current threshold is set to 1.9. Why?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Wavelet analysis for fault diagnosis is a very old technique. There is no innovative technology in the work. And actual testing is lacking. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf