SRMANet: Toward an Interpretable Neural Network with Multi-Attention Mechanism for Gearbox Fault Diagnosis

Round 1

Reviewer 1 Report

The authors proposed Stacked Residual Multi-Attention Network (SRMANet) for feature extraction of vibration signals and provide a study in detail. The results look encouraging and motivating. But there are still some contents, which need be revised in order to meet the requirements of publish. A number of concerns listed as follows:

(1)   The abstract should be narrow down on the problem and highlight the need of the proposed work with experimental results. Add the contents in the abstract of the paper and highlight the impact of the proposed work.

(2)   In the introduction section, you should give the novelty and the contributions of your works.

(3)   The methodology is not clear and it can be further improved it is better to add a flow chart of methodology.

(4)   The method/approach in the context of the proposed work should be written in detail.

(5)   The literature review is poor in this paper. you must review all significant similar works that have been done. For example, 10.1016/j.isatra.2021.07.017; 10.3390/agriculture12060793; 10.1109/JSTARS.2021.3059451;10.1007/s10489-022-03719-6 and so on.

(6)   Line 200, the expression is not clear, please revise it.

(7)   Line 225, Un mn is ?

(8)   Result and discussion should be rewritten to summarize the findings/significance of the work.

(9)   The paper is in need of revision in terms of eliminating grammatical errors, and improving clarity and readability.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper is concerned with a fault diagnosis approach applied to gearboxes, which is a hot research topic and a very important one when field applications are considered. A suitable survey of published material related with the paper’s topic is included in the References section and appropriately mentioned in the text. Considering several gearbox faults, several experiments were conducted to assess the robustness of the proposed approach and the obtained results were critical analysed and compared with results of other reported methodologies. However, the paper to be recommendable for publication needs major revisions according to the following comments:

-          Throughout the paper, there are several missed spaces between the punctuation and the following words;

-          It is not clear how the scaling factor Ɛ is determined;

-          It should be clarifies for what stands the parameter γ in equation (3) and (5), as well as the parameter b in equation (4);

-          The description of the nomenclature used in equation (7) should be revised (sentence following the mentioned equation). There is an index m associated with the variable u_n, which doesn’t appear in the equation. Similar situations happen with equations (8) and (12);

-          It seems that something is missed in equation (10);

-          The way figures are mentioned in the text should be normalized by using the same format for all of them;

-          Algorithm 1 should fit within the page without considering the document margins. Furthermore, it should be used the same format for all text;

-          Figures captions should be located in the page where the corresponding figure is placed;

-          A line space should be given between Table 1 and the following paragraph;

-          Figure 5 should not be splitted between pages and should fit within the page limits without considering the document margins;

-          Tables 3 and 4 should not be splitted between pages;

-          Figures 7, 8, 9, 10, 11, 12 and 13 should fit within the page limits without considering the document margins;

-          It should be deeper described how the parameters included in Table 3 were tuned;

-          References section should fit within the page limits.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

ccording to the revised paper, I have appreciated the deep revision of the contents and the present form of this manuscript.  There is little content, which need be revised according to the comment of reviewer in order to meet the requirements of publish. A number of concerns listed as follows:

 (1) The authors need to interpret the meanings of the variables.

 (2) Please highlight your contributions in introduction.

 (3) How to determine these parameters? The author should give a detailed explanation.

 (4) Conclusion: What are the advantages and disadvantages of this study compared to the existing studies in this area?

 

(5) The inspiration of your work must further be highlighted. Some suggested recent literatures should add in the revised paper according to the previous comments of reviewer .

 

(6) Further correct typological mistakes and mathematical errors.

 

I hope that the authors can carefully and further revise this manuscript according to the reviewer comments in order to meet the requirements of publish.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx