Multi-Objective Task Scheduling of Circuit Repair
Round 1
Reviewer 1 Report
1. Include some quantitative results in the abstract if it’s available.
2. Carefully define all the acronyms in the abstract and body.
3. The fonts in Fig. 1 is too small, so the figure should be in landscape for clarity.
4. There are some repetitive statements in the introduction.
5. The quality of Fig. 5 must b improved.
6. What's the novelty of this work.
7. Overall, this is a good work with a conceptual clarity.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This paper provides three rest solutions and analyzes flexible rest scheduling for repair teams. Its authors built a scientific, complete mathematical model for circuit repair scheduling. This model maximizes advantages and minimizes risks while scheduling up to a certain point given restrictions like geographic information and resources. Moreover, they utilized NSGAII-2Opt-DE(N2D), SPEA2-2Opt-DE(S2D), and MOEA/D-2Opt-DE (M2D) to determine which is the superior algorithm or technique that will solve their problem. Also, they develop an area indicator. It helps assess multi-objective optimization method convergence speed.
The authors stated that they do not have the ability to obtain comprehensive data. Therefore, 10 sets of data that are randomly generated are used in this article. Since they are submitting this paper to this journal, they should use comprehensive data and not randomly generated datasets. How can we be sure that this randomly generated data be applicable if applied in the real-case scenario?
Author Response
- The authors stated that they do not have the ability to obtain comprehensive data. Therefore, 10 sets of data that are randomly generated are used in this article. Since they are submitting this paper to this journal, they should use comprehensive data and not randomly generated datasets. How can we be sure that this randomly generated data be applicable if applied in the real-case scenario?
First, the section “5.4. Algorithm Scientificity” cites references [42-44], which provide theoretical proofs for the effectiveness of the algorithms.
Second, there is less targeted literature available. For the first time, we propose parameters such as “risk” during the transfer of repair teams. Both the relevant literature and the datasets available on the Internet have the problem of incomplete data collection. So we generate the data randomly. In this article, we generate data according to uniform distribution and study various cases with equal probability. Real data is a special case of random data. If there is an order of magnitude difference between the real data and the data in this paper, please use normalization. We conducted 600 experiments and obtained good results. To a certain extent, the scientific validity of the model and algorithms proposed in this article are verified through experiments.
Therefore, through theoretical and experimental proof, we ensure that these randomly generated data are applicable to the real-case scenario.
We greatly appreciate your valuable comment. If you need anything else, please let us know and we will further revise and improve.
Reviewer 3 Report
The authors propose a multi-objective task scheduling of circuit repair approach.
The approach prosed is very interesting and the contributions are highlighted.
Some suggestions to improve the paper:
Add more arguments on the use of NSGAII-2Opt-DE(N2D), SPEA2-2Opt-DE(S2D), and 16 MOEA/D-2Opt-DE(M2D).
More explanations on how the parameters of the algorithms have been fixed.
More comments on constraint handling are needed.
The literature review needs to be polished by some recent references.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
I may agree with the authors about generating data according to a uniform distribution, studying various cases with equal probability, and justifying the use of normalization. Moreover, I did not see in their manuscript that 600 experiments were conducted; I only saw 600 iterations. Are they similar? In addition, 600 experiments using computer simulations are not enough. Usually, we use more than 1000 experiments. Then, plot the Bell curve to see that the experiments made are close to the correct figure or specific value. I want to see the analysis made when 600 experiments are conducted, but I prefer 1000 experiments to be conducted.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
The authors have already revised their paper based on some of my comments. Moreover, I am satisfied with their arguments for some of my suggestions for their paper revision.
