Intelligent Optimization Design of Distillation Columns Using Surrogate Models Based on GA-BP
Round 1
Reviewer 1 Report
1. The introduction section should be reorganized. The existing version is not clear.
2. In section 2, authors simply list the method of literature. It is suggested to revise this section with the advantage and disadvantage of these methods clarified and summarized.
3. In Section 3, only the optimization procedure is introduced, while the detailed information of the model is not provided. Please revise.
4. Compare your results with others from the literature and discuss them.
It is highly recommended that the authors thoroughly check grammar and spelling errors to improve the overall clarity of the paper.
Author Response
Dear reviewer:
Thank you for your decision and constructive comments on our manuscript. We have carefully considered your suggestion and made some changes. Our responses are given in a point-by-point manner below. Changes to the manuscript are shown in underline/red/bold.
Point 1: The introduction section should be reorganized. The existing version is not clear.
Response 1: Thank you for your suggestions in the introduction section. We have re-written the introduction section.
Point 2: In section 2, authors simply list the method of literature. It is suggested to revise this section with the advantage and disadvantage of these methods clarified and summarized.
Response 2: Based on your suggestion, we have revised Section 2 to clearly summarize the advantages and disadvantages of these methods, as shown in Table 1 and Table 2.
Point 3: In Section 3, only the optimization procedure is introduced, while the detailed information of the model is not provided. Please revise.
Response 3: In order to illustrate the model building process more clearly, we have added explainations between line 262 and line 283.
Point 4: Compare your results with others from the literature and discuss them.
Response 4: We have added a detailed comparison and discussion of the design results of this work with the results of others from line 556 on page 19.
Finally, we have thoroughly checked the manuscript for grammatical and spelling errors and have made some changes marked in red in the revised manuscript that do not affect the content or framework of the paper. We would like to express our sincere gratitude to the reviewers for their enthusiastic work and hope that the revisions will be recognized. Thank you again for your comments and suggestions.
Reviewer 2 Report
The manuscript suggests an intelligent optimization design method of distillation columns by surrogate model based on GA-BP. The algorithm is sound and applicable. It is helpful for the optimization design of such kind of equipment. The manuscript should be improved in the following aspects.
1. The terminology should be consistent through the paper. For example, reflux rate and reflow rate.
2. Take care of abbreviations. Generally, the full expression should be given at the first time the abbreviation appears. Check abbreviation MESH, CDU, etc. The letter used in the abbreviation should be capitalized in the full expression correspondingly.
3. The variables should be more meaningful. For example, Rc is better than RR for the calculated reflux ratio.
4. Try to use a single letter for a variable, but with subscript or superscript for additional meaning. For example, Fe may be used for the emission factors.
5. Check equation (1) and the following text. Is there any p in the equation?
6. Check equation (3) (4) and the following text. Is there any N in the equations?
7. Check equation (6) and the following text. What is the difference between the two symbols for density? Is the subscript still misspelled for steel?
8. In Fig.5(b), line types with obvious difference should be preferred for best and goal. In Fig.5(c), the width of line for Fit should be moderate.
9. In Fig.1, links between the "select, crossover and mutation" and the BP-NN should be added.
10. The content from line 268 to line 288 should be re-arranged to be more logic.
11. In Equation (13), min should also be added before function symbol f.
12. In Equation (13), if N=N1+N2, N1 and N2 are not independent variables, N1 or N2 should be cancelled. The optimization result should include all design variables.
13. The optimization model should be more complete with constraints, not only design variable boundaries, but also state variables boundaries.
14. Constraints check in Fig.1 should be explained in corresponding context. GA generally changes constraint optimization problem to unconstraint one by some kind of penalty function. Is there any special treatment?
1. Read the manuscript carefully, there are some confusing expressions. For example, in line 230, what is the meaning of "such as the numbers of N, Nf etc."?
2. Read the manuscript carefully, there are some missing punctuation marks. For example, in line 474, full stop should be added after "...in Table 4".
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 3 Report
The paper is a nice piece of Intelligent Optimization Design of Distillation Columns Using 2 Surrogate Models Based on GA-BP. The reviewer suggests revising the following items for possible improvement of the work.
1. Graphical abstract indicates interconnection/interface between Aspen Plus and MATLAB. How to direct link between Aspen Plus and MATLAB? Please provide the general information.
2. “The sample data is obtained through rigorous simulation, considering the design and operating variables.”. How to determine the initial point of design and operating variable in the rigorous simulation to generate the data for optimization if authors don’t use previous published data?
3. The proposed method is implemented in the simple case of Propane and Propylene which can be easily optimized. Due to sample data generation also requiring time, is it possible for the authors to clarify how efficient the proposed method is for simple cases compared to the common design and optimization (shortcutàrigorousàiterative optimization or rigorous automatic optimization by linking Aspen and MATLAB).
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
All comments are addressed.
