CNN-Based Feature Fusion Motor Fault Diagnosis

Round 1

Reviewer 1 Report

A multi-feature extraction module based on attention mechanism is proposed in this paper by the authors. . The experiment is carried out to prove that the motor fault diagnosis method is of high diagnosis accuracy rate. The paper is well written and contribute a novel idea for fault detection. However, it requires some minor revision:

1. Minor English corrections are required. Mainly in abstract section.

2. For result only confusion matrix and  accuracy are added. What about the other performance measures of classification? 

Author Response

Dear reviewer,

Thank you for your valuable comments. They are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have tried our best to revise our manuscript according to the comments. The main corrections in the paper and the responds to your comments as shown in the attachment, Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper’s subject is relevant and interesting. Authors showed well results in the signal classification. The study is illustrated by a well-experimental investigation

 

Recommendations and comments

1.    The Introduction should be increased by analysis of methods in non-destructive diagnosis and testing. The presented version of the introduction is dominantly focused on the analysis of CNN. Can other methods of machine learning be used in motor fault diagnosis? Could you compare your result with the methods proposed in:

Rabcan, J., Levashenko, V., Zaitseva, E., Kvassay, M., Subbotin, S. Non-destructive diagnostic of aircraft engine blades by Fuzzy Decision Tree, Engineering Structures, 2019, 197, 109396

Zhu, Y., Li, G., et al. Intelligent fault diagnosis of hydraulic piston pump combining improved LeNet-5 and PSO hyperparameter optimization, Applied Acoustics, 183, 2021,108336

Glowacz, A.,Fault diagnosis of electric impact drills using thermal imaging, Measurement: Journal of the International Measurement Confederation, 171, 2021, 108815

2.    Why is the accuracy (Table 5) used for the result evaluation? What can you say about Sensitivity, DOR, F1, Youden’s index, Jaccard index, and others?

3.    Could you use other methods expected CNN in comparison of your result?

4.    Could you explain the novelty of your result more clear?

Author Response

Dear reviewer,

Thank you for your valuable comments. They are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have tried our best to revise our manuscript according to the comments. The main corrections in the paper and the responds to your comments as shown in the attachment, Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Authors considered my comments. I have not other recommendations.