Energy Efficiency Optimization in Onboard SWRO Desalination Plants Based on a Genetic Neuro-Fuzzy System

Round 1

Reviewer 1 Report

The paper presents the use of a Genetic – ML approach for the determination of optimal target values for the control inputs of a desalination plant.

Although interesting, with an original application, the paper is difficult to read, and I think that it is particular hard to understand the different steps of the methodology. Many paragraphs should be rephrased to enhance the understanding of the paper.

In a particular :

-          Fig.1 provides absolutely no information. What ere the inputs, the outputs, the variables, the components (E.A ? F.C ? B.A.P. ???)

-          Why does the 24 hours horizon, with a sampling rate of 1 minute, ends with sample no. 1149?

-          I do not understand what the « cost function represents ». In ters of requirement, we satisfy the constraints or not, we cannot express the requirement as a cost function. After that, if it is difficult ti ssatisfy those requirements, we can express them by means of penalty functions, with different weights. Is it what you are doing when you talk about 50%, 30%, 20% ?

-          In that case, an equation would be better: are 50%, 30%, 20% the weighting factors, or do you normalize before the 3 differents terms with respect to nominal values so that a percentage makes sense?

-          If you have a cost function, you can evaluate the cost for all samples. How do you decide to select the best 828 samples? Do you have in addition some hard constraits to satisfy?

-          The description of your three layer is really hard to follow. A figure describing the organization of these layers together with the used variables could help.

-          I have the feeling that I the end, you have a regression technique to fit a relation between static input/output data. In that case, some comparison with other standard methods should be given.

Author Response

We attach the answers in the document.

Author Response File: Author Response.pdf

Reviewer 2 Report

Suggestions to improve the paper:

1). Reference part need to be strengthened.

2). A flow chart of the step - wise methodology should add value to the paper.

3). Minor language editing may be required. 

 

 

 

Author Response

We attach the answers in the document.

Author Response File: Author Response.pdf

Reviewer 3 Report

The method applied is explained too much in theoretical way and it is not explained how the GNF algorithm has been developed (i.e. which calculation tool was used, how many interactions, ...) 

The final results obained are good, but if someone wants to reproduce them, it is quite impossible.

Author Response

We attach the answers in the document.

Author Response File: Author Response.pdf

Reviewer 4 Report

The manuscript is clear, relevant for the field and presented in a well-structured manner.

The cited references are mostly recent publications (within the last 5 years) and relevant. It doesn't include an excessive number of self-citations.

The manuscript is scientifically sound and the experimental design appropriate is to test the hypothesis.

 

However, there are several questions and comments for the authors of the manuscript.

1. What type of membership functions is used for fuzzy systems?

2. Please provide the architecture and structure created by Genetic Neuro-Fuzzy in the manuscript. How many elements are in the hidden layers?

3. Tables 3 and 4 should be merged into one.

4. Figures 3, 4 and 5 are not informative.

5. Give examples of fuzzy rules inference used in the system.

6. How many fuzzy inference rules were used in the system?

7. What type of fuzzy systems are used in the manuscript?

Author Response

We attach the answers in the document.

Author Response File: Author Response.pdf

Reviewer 5 Report

 

In the summary and introduction, the explanation should be given where SWRO is first mentioned.

At the end of the abstract section, it should include the numerical (or tangible) results of this study.

I didn't quite understand the novelty of working on Line 72. Please describe the novelty of the study in detail.

In addition, the contributions of the study should be explained in the introduction.

Why are all parameters 0 in Figure 1?

Why did you choose the genetic algorithm for the study?

There are typos in the text in general. A careful syllable check should be done in the text.

The figures' quality is not good and should be improved.

References should be in accordance with the journal format.

I understand that one of the aims of the study is to increase efficiency. Is there any increase in efficiency?

The results and discussion part of the study should be revised and explained in detail.

The literature search of the study is weak.

 

 

Author Response

We attach the answers in the document.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The journal allows so little time to review the paper that it is difficult to provide a fair and detailed review.

However, at first and quick sight, it seems that the authors have tried to answer my remarks.

Author Response

Thank you very much for the comments. We are happy we have answered the questions about the article.

 

Reviewer 5 Report

If possible, the literature should be checked again and the literature part in the introduction should be explained in more detail.

The article should be checked for typos and language grammar.

 

Author Response

Thank you again for the comments, we appreciate it. We have read carefully the article and we have corrected small typos and language grammar mistakes. On the other hand, we have added some sentences in the introduction of the paper to explain the literature in more detail.