Wind Power Converter Fault Diagnosis Using Reduced Kernel PCA-Based BiLSTM

Round 1

Reviewer 1 Report

1.Figure 2 is missing some of the variables in the calculation formula, and if added, you can better show the relationship between each variable and the output.

2. I think that if the results of the preprocessing of the previous data are increased, the processing effect of the  Euclidean Distance and the Hierarchical Kmeans clustering in the early stage will be more convincing.

3.The data processing results obtained by the experimental method in this paper are only presented in tabular form, and there is a lack of stronger evidence.

Author Response

We are grateful to Reviewer 1 for his/her valuable comments. We do appreciate
the time spent on the detailed comments. The comments are addressed below and
are incorporated in the revised manuscript. We believe that the manuscript has
been improved considerably while addressing the reviewer comments.

Author Response File: Author Response.pdf

Reviewer 2 Report

Abstract: would you put "(fault detection and diagnosis)" right after FDD?

 

equation 11: sigma is used for the sigmoid function in eqn (1), may use another symbol to avoid confusion

 

Figure 6: Please identify the meanings of Fault3 and Fault9, do you mean x3 and x9?

 

For the operation of the WTC system samples, do you use five-fold cross-validation?

 

Table 9, would you simply explain why type I and type II errors (the other diagonal) in the confusion matrix are zero in your case?  Will type I and type II errors be avoided entirely?

 

The fifth line in the Conclusion section: change "build The reduced" to "build the reduced".

Author Response

We are grateful to Reviewer 2 for his/her valuable comments. We do appreciate
the time spent on the detailed comments. The comments are addressed below and
are incorporated in the revised manuscript. We believe that the manuscript has
been improved considerably while addressing the reviewer comments.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper presents a method to reduce the complexity and computation time and increase the accuracy of fault diagnosis in wind power systems.

However, the following comments need to be considered to improve the quality of the manuscript:

1) In the abstract. what is the power system used to test the proposed method for fault detection? What type of faults considered in this work? It is not mentioned.

2) In the abstract, what is the improvement done to the studied system? In percentage, accuracy,...etc

3) A nomenclature table should be provided in the beginning of the paper.

4) The Wind Energy System (provided in Section 3) should be added in the modelling Section 2.

5) Where are the parameters values taken from (Table 1)? Are they arbitrary or taken from a reference? Please specify.

6) Some minor mistakes need to be revised, as catastrophic (section 3.2) and Mixed in Table 2.

7) In section 3.3, it is written "Hkmeans". Is it a typo? (K-means)? What are the differences between K-means and HKmeans?

8) What is the simulation platform used for this work? Nothing is mentioned about it in the manuscript. Which platform is used to code the KPCA-BiLSTM, ANN, MNN, CFNN,.....etc. Are the ANN, MNN,..in Table 5 from previous manuscripts in the literature? Or where they built by the authors for comparison purposes?

9) Is the work done in this paper only compatible with Wind Energy Systems? Or it can work with other larger and more complex systems? What are the challenges in bigger systems?

Author Response

We are grateful to Reviewer 3 for his/her valuable comments. We do appreciate
the time spent on the detailed comments. The comments are addressed below and
are incorporated in the revised manuscript. We believe that the manuscript has
been improved considerably while addressing the reviewer comments.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The authors addressed all comments raised in the previous round. Therefore I recommend the acceptance of the paper in its present form.