Nonlinear Dynamic Process Monitoring Based on Two-Step Dynamic Local Kernel Principal Component Analysis

Round 1

Reviewer 1 Report

line 40: Hajer et al. proposed a  ... 

 should be: Lahdhiri et al. proposed a ...

in the lines 53, 120, 128, 134, 151: is  the expectation  ...

should be: ...the expectation values of ...

line 177: What would the end result be if a variance value other than 0.01 was used? 

 

Author Response

All replies are attached.

Author Response File: Author Response.pdf

Reviewer 2 Report

Please see attached review report

Comments for author File: Comments.pdf

Author Response

All replies are attached.

Author Response File: Author Response.pdf

Reviewer 3 Report

Report on the manuscript "processes-1663049" entitled "Process Monitoring of Nonlinear Dynamic Processes Based on Two-Step Dynamic Local Kernel Principal Component Analysis"

This manuscript proposes a new two-step dynamic local kernel principal component analysis that extracts the static components in the process data and then analyzes them with this analysis. Experimental results are provided. Conclusions about the present investigation are reported.

I have a good opinion about this work, but some issues must be fixed. The following are my comments regarding this manuscript:

1. The manuscript needs to be proofread by the authors. I have noted some drafting problems.

2. The title of the manuscript must be improved: "Process" Monitoring of Nonlinear Dynamic "Processes". The authors must remove one of the words "Process".

3. Words in the title are not usually in the keywords. In addition, the keywords are often written in alphabetical order. Please do not define abbreviations in the keywords.

4. The authors must check the use of all acronyms, abbreviations, and notations employed in the whole manuscript. I have noted a lot of typos and bad use of notations. In addition, some notations must be written in text style and not in math style; see, e.g., Eq. (7). Please check it in the whole paper. I suggest using "T" for the transpose of vectors and matrices instead of "T" in Italic font to avoid confusion with math variables.

5. Some sections start with comments and others do not. The authors must do an effort to improve the presentation and organization of their manuscript, which will help its readability.

6. Is this manuscript an application of known methods to a real problem or a new methodology is developed? This must be clearly stated. 

7. The authors must provide more details about the computational framework used in the manuscript. For example, software and packages used, features of the computer employed, runtimes, and other computational aspects must be added. 

8. The authors could summarize their methodology into an algorithm and in a flowchart so that the readers can follow it easier. Thus, the practitioners could have some guidelines when applying this methodology.

9. I do not have each numerical result in detail. I recommend the authors to check them.

10. In my opinion, the implications and results of the study are underdeveloped and must be improved and explained further in the final section.

11. The conclusions need to be improved. Also, the authors must add limitations of the study and ideas for further research. Then, I suggest titling the final section "Conclusions, limitations, and future research".

12. The authors must check whether all references are cited and whether all citations are in the reference list. The authors should do an effort to cite papers on the topic published in Processes to attract the attention of our target audience. The format for references must be improved.

13. The bibliographical review must be improved. I suggest the authors present a table that shows the state of the art with the advances on the topic. The references

https://doi.org/10.3390/math10060890

related to the topic can be followed.

Author Response

All replies are attached.

Author Response File: Author Response.pdf

Reviewer 4 Report

Statistical Process Control (SPC) methods, such as, e.g., control charts, were introduced for monitoring discrete production processes. Recent developments in this area include the usage of control charts for monitoring multivariate quality characteristics. However, these classical methods are insufficient for monitoring many other processes, such as continuous processes of the chemical industry. For monitoring such processes, new methods have been proposed that use advanced methods of statistics and data mining. These methods require different than those used in SPC, criteria for the evaluation of their efficiency. The reviewed paper proposes a new method for monitoring such complex processes. It can be regarded as the extension of already proposed methods, including those proposed by the coauthors of the reviewed paper. Therefore, the reviewed paper contains original results whose quality has been demonstrated using simulation methods and other methods used in the chemical industry, such as the Tennessee Eastman Process test.

Before its recommended publication, the paper needs a thorough editorial process. Language-wise errors must be corrected, all used symbols have to be described, and the list of references MUST be improved (e.g., some names of Journals are lacking!).

Author Response

All replies are attached.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Manuscript may be accepted in present form

Author Response

Thanks

Reviewer 3 Report

I have checked the authors' revision and I am satisfied with the new version of the paper. Therefore, I recommend its acceptance in the present form.

Author Response

Thanks.