A Fault Diagnosis Approach for Electromechanical Actuators with Simulating Model under Small Experimental Data Sample Condition
Round 1
Reviewer 1 Report
The paper sounds. However some minor improvements are recommended to be performed, namely the simulation models description shall be strengthen in terms of research reproduction by other researchers.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
1) In such important work with a wide field, it is necessary to approach several articles in the field in the introduction, and at the same time to broaden the reference base.
2) The citations must be given in order, not mixed, so references 25-26 must be removed and put at the final, in order to be ordered.
3) What's the point of line 127 "Error! Reference source not found."?
4) Figure 3 is unclear and requires increased resolution.
5) Table 1 and Figure 5 require additional local discussions and comments.
6) Figures 6, 7, 8 also require a broader local discussion.
7) Paragraph 4.2 discusses both the training method needed to diagnose the fault and the transmission through the network using neural methods, but the problems are slightly mixed and reference is made for understanding to the paper [26]. The authors have a duty to explain more clearly the two distinct phenomena, the training for fault detection, and the use of neural algorithms for both measurement, detection, control, and the use of a network. More details are desirable here and some clarification discussions.
8) Before "Conclusions", an additional section should be introduced, "Discussion", in which to be highlighted all the new methodology proposed in the paper, the new results obtained, compared to other existing methods, to be highlighted the advantages of the new method and its limitations. It was stated in the paper that the proposed method has high efficiency, but here in the discussion chapter, it will be highlighted compared to other existing methods that the more efficient the proposed new method is. Also here it would be good to propose new possible future directions of work opened by the method proposed in the paper. Some of the limitations of the methodology were discussed in the conclusions, but there it is better to specify more what is new and good in the paper, advantages, and perspectives, while the limitations are discussed in the additional chapter "Discussion".
9) It is desirable that the number of references to be expanded.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The contribution of this work is very good and deserves publication. The title has been formulated unambiguously conveying the focus of the study.
Appropriate research goals are chosen in this contribution, which shows that the authors have a high level of understanding of current research within the field of their research. The authors successfully used the appropriate techniques for analysis of the research objects.
The accurate interpretation of outcomes, well substantiated by the results of the analysis has been achieved by them. The presentation of the results in terms of the research objectives has been successfully made.
The authors have been able to draw logical conclusions from the results. The results and their respective discussion points prove the efficacy of the proposed method. Conclusions are accurate and clearly based on outcomes.
In Figure 2. EMA Simulator shows the control scheme in a cascaded form.
Please analyze the stability of the scheme more in depth such that the Reader can understand better the contribution and can reproduce the results.
In Figure 15. a Fault diagnosis model Architecture is shown. could you please explain better the scheme and discuss if some singular conditions can happen in which a stand still situation can occur?
In fact, it is not clear if there are fedback in this structure.
Please clarify!
The contribution of this work is very good and deserves publication.
Concerning the cited literature you can consider the following paper to improve the tutorial aspects of the paper.
Paper in MDPI:
Mercorelli, P. A Fault Detection and Data Reconciliation Algorithm in Technical Processes with the Help of Haar Wavelets Packets. Algorithms 2017, 10, 13. https://doi.org/10.3390/a10010013
Villani M, Tursini M, Fabri G, et al. Electromechanical Actuator for Helicopter Rotor Damper Application. IEEE Transactions on Industry Applications 2014;50(2):1007-1014. 440
Rauf H T, Bangyal W, Lali M I. An adaptive hybrid differential evolution algorithm for continuous optimization and classification problems. Neural Computing and Applications 2021;(33):10841-10867.
Author Response
Please see the attachment.
Author Response File: 
Author Response.doc
Reviewer 4 Report
This paper has made a certain contribution to the field of research on the fault diagnosis of electromechanical actuators. Generally speaking, the paper is well structured, with sufficient details of the simulation and experiment. However, the following issues need to be improved before publication:
- Due to the lack of discussion of other research in the field, the introduction is not detailed enough.
- Training data in deep-learning should be supported more strongly to verify the reliability of input.
- In order to show the effectiveness of the experimental accuracy, it is suggested to add the description of the measures to resist the interference of external factors in the experiment.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
1) The last table should be numbered 3 instead of 1.
In the current (revised) form the paper is suitable for production.
