Multi-Objective Optimization Design of Permanent Magnet Eddy Current Coupler Based on SCG-BP Neural Network Modeling and the ONDX-NSGA-II Algorithm
Round 1
Reviewer 1 Report (Previous Reviewer 2)
I have reviewed the previous version of the paper and asked a question, which authors cannot ignore. However, they did.
Line 295 defines the relative slip of 40.58 rpm or 4.25 rad/s
Transfered torque is 4.1015 kH*m.
The total losses in the system are slip * troque = 17430 W which is smaller than eddy current loss 19303 W in Table 2.
This cannot be true!!! It is against law of energy conservation...
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report (New Reviewer)
This paper dealing with the multi-objective optimization of a PMECC device used in transmission is in its second revision stage. Still, more elements need to be discussed/corrected.
Issues/comments to correct/discuss:
- In the introduction section the use of PMECC is not justified against other solutions (like the mechanical gears and magnetic gears). Why is it more appropriate/efficient to use eddy current coupler devices? A state of the art on this topic would be necessary to justify the solution (it could be in terms of torque density, efficiency, cost, reliability). In this context, the authors should clearly indicate their focused application (the possible applications indicated in the first paragraph are opening a too wide range of destinations for the PMECC).
- The idea considering the toque among the optimization parameters is not justified, together with the eddy current and the cost (of the Active Part of the PMECC, I suppose), for two reasons. First, optimization means minimization of something, like the cost and the eddy current loss, and, on the other hand, the torque shouldn’t be minimized. Secondly, we cannot speak of torque maximization since the value of the torque is imposed by application. It doesn’t make sense to speak about torque increase, instead it makes sense to talk about torque-density increase (by optimization, the volume of the structure is minimized while the desired torque is obtained).
- I don’t see a logic justification for the multiplication of the volume of the magnets by 10 (according to equation 7 and its related text). The volume and the cost per kg shouldn’t be altered by any coefficients…
- In the paper it is not indicated (plotted) how the optimized structural parameters evolved in the optimization process. Plots (2D) of all parameters indicated in Table 1 should be added to the paper. By the way, the margins of the structural parameters which are establishing the space of solution should be given into the paper (as well as other possible constraints).
- The multi-objective parameters have varied during optimization process and 3D plots are presented in Fig.8-9. A 2D representation of each parameter (torque density, eddy current and cost) needs to be depicted in the paper for the reader to have a clue about the considered optimized solution.
- Normally, for such high thickness magnets (2.5cm) and torque density the PM losses cannot be neglected. Also, the paper doesn’t specify anything about the efficiency of the PMECC; this information should be given, for the original topology as well as for the optimized structure.
- The eddy current loss level has increased on the optimized structure (according to fig.13). So, how efficient is this hybrid optimization process?
- On the numerical and experimental plots, which is the rated operating point?
- The efficiency of the PMECC can be experimentally indicated, isn’t it? By using the load and load conditions the mechanical component loss of the PEMCC can be obtained. Thus, the experimental part of the paper will be more consistent.
No suggestions about English.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report (New Reviewer)
This is a very good paper about the design and application of a new optimization algorithm. It can almost be accepted in present form. You should have a look at the units, they must not be written italic.
English is good. Somebody should look over the spelling, e.g. page 3 line 95: why "Control" in capital letter?
Author Response
请参阅附件。
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report (Previous Reviewer 2)
Authors addressed all my previous questions.
I suggest to accept the paper.
Author Response
Dear reviewer, thank you for your careful review and valuable feedback on our manuscript. We are very pleased that you have been satisfied with our responses to your previous questions.
Thank you again for your time and professional review.
Reviewer 2 Report (New Reviewer)
The reviewer appreciates the authors' care for responding to all the addressed issues. I would anyway suggest that part of the explanations in the cover letter could be added to the paper (for example the reason for not showing the structural parameters variation during the optimization process, the reason for not showing the PM loss component, the reason for discussing on torque maximization instead of torque density maximization… the paper would be increased with supplementary explanations. Even if the comparison with magnetic gears was not treated in the introduction section, the paper has significantly improved.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
The authors proposes a multi-objective method for optimization of permanent magnet eddy current couplers. The optimization method is based on the hybrid approach. It combines together SCG-BP neural network modelling and improved ONDX-NSGA algorithm. The proposed method of optimization of permanent magnet eddy current couplers has better performance than traditional methods. The conclusions are consistent with the evidence and arguments presented. All the references are appropriate and cited correctly in the manuscript. Additionally, there are no any excessive self-citations.
Some comments to improve quality of the paper:
1) Experimental verification is still recommended in order for the paper to be more interesting and valuable for readers and researchers.
2) Table 2: What is unit of measurement of the cost?
3) Abbreviations SCG-BP and ONDX-NSGA should be expanded once in text.
4) Figure 9: titles of axes should be shown along the axes.
Minor editing of English language required
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
No every abbreviation is defined (SCG-BP, ONDX-NSGA-II, NDX).
Do not use abbreviations in the Abstract
Please show the power balance diagram. Some of the numbers are definitely wrong.
For example, let's take a look at Table 2.
The system is working at the slip of 40.58 rpm or 4.2495 rad/s
The produced torque after optimization is 4.4517 kNm
It is known that the losses are (this equation is derived from the law of energy conversion and they are minimal possible losses): Ploss = slip * torque = 18.918 kW
This value is bigger than indicated in Table 2 (18.586 kW), and that is impossible.
Again, before optimization the power of losses is 17.429 kW which is smaller than in the Table 2.
So, the point after optimization is incorrect, and for point before optimization it is necessary to draw power diagram to show the power flows and to explain these additional losses.
Again, in Table 3 "Finite element" losses should be at least 19 kW, while they are 18.6 kW in Table 3, which cannot be true.
I see that research is not conducted correctly. Some of the results contradict the law of energy conversion. I suggest to reject the paper with resubmission after major revision.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper describes the multi-objective optimization technique for the permanent magnet eddy current coupler. A neural network is used to accelerate the evaluation the device performance. The paper can be published. However, some questions should be answered before the publication.
1. The field of applications of the permanent magnet eddy current coupler is not quite clear. Eddy currents produce losses. In which applications is the device can not be replaced by the coupler with permanent magnets on both parts?
2. How was number of poles chosen? Why was it not optimized?
Also, please, proofread the paper. For example, ‘silp’ in Fig.2.
Author Response
Please see attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
I was not precise enough when mentioning loss equation.
We have input power P1 = T * omega1
We have output power P2 = T * omega2
The losses are: deltaP = T * (omega1 - omega2).
This is the precise equation.
The modelling experiment in Table 2 was conducted with the slip of 40.58 rpm or 4.2495 rad/s.
If we multiply T by slip we get:
17.43 kW of total losses for "before optimization" with the torque of 4.1015 kNm. The eddy current loss are stated to be 19.3 kW, which is bigger than total losses in the system. Thus, at least this point is wrong.
If this point is wrong, how can I trust other results in the paper?
