Optimization of Neural Network-Based Self-Tuning PID Controllers for Second Order Mechanical Systems

Round 1

Reviewer 1 Report

. The concept presented is good and the effort of the authors is well appreciated. . The result is not clear, needs to be improved and reorganize.

Author Response

Response to the reviewer comments is included in an attached file (Cover_letter_and_Response_to_reviewer_.docx). We kindly request that you open the document to view the responses.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear authors,

This paper proposes a method to identify the model parameters and PID controller coefficients using neural networks. The text and structure of the paper are satisfactory, however, there are some major issues that should be addressed to improve the paper:

1. The abstract should be more informative about the innovation of the paper. The structure of the proposed method is not clearly described in the abstract. Please avoid writing general phrases such as the first sentence.
2. Fuzzy logic and adaptive neuro-fuzzy inference system (ANFIS) were also applied to tune the PID controllers in the literature. What are the advantages of your method over them? Please describe them in the Introduction.
3. Subsection 2.2, second paragraph: Where is subsection A?
4. Subsection 2.3, second paragraph: “The ANN received inputs that included n sampling with values at 0.1 intervals…”; It is unclear, samples form which signal?
5. Subsection 2.3, third paragraph: “The neural network inputs are given via four approaches”; What are these approaches? Please describe them in the text.
6. Explain the “system characteristics” in a Table.
7. Subsection 2.4, second paragraph: Please provide a formula to explain the added noise.
8. Figure 5: “Acceptable response?”, It is unclear which response is acceptable? Which parameters have been used to decide about the system acceptability?
9. Show some examples to describe the performance of the first neural network for model identification in the Results section.
10. Table 2: Why the performance of the L-21 (no response characteristics) is better than L-21-RC? Please mention that you used 10 million data in the caption.
11. Which algorithms were used to train the neural networks? How was the overfitting issue prevented?
12. Results Section, seventh paragraph: “It is suggested to evaluate the situation in which there is noise with a certain margin of evaluation criteria.” This sentence is vague. Please elaborate more on the robustness of the system in the presence of noise.
13. The last paragraph of the conclusion should be deleted. It is part of the template text.

Best regards

Author Response

Response to the reviewer comments is included in an attached file (Cover_letter_and_Response_to_reviewer_.docx). We kindly request that you open the document to view the responses.

Author Response File: Author Response.docx

Reviewer 3 Report

The thread of this paper is clear and the expression is accurate. However, there are some weakness：

1. There is something wrong with the format of the article，such as the line146， 227 and so on.
2. The intervals in Figure 3 are not well-distributed，which conflicts with the definition in the text.
3. Please reinterprete the situation when the target position is changed and the Figure 8. It's a little hard to understand.
4. What is the meaning of changing target position? Is it better to plot the figure of response instead of the figure of position?
5. Some latest litratures on modifed PID controllers should be included in Introduction, see "Quantitative Tuning of Active Disturbance Rejection Controller for FOPDT Model with Application to Power Plant Control. IEEE Transactions on Industrial Electronics" 
6. What‘re the differences between the experiment of noise and the experiment of changing position?
7. Language should be polished. 

Author Response

Response to the reviewer comments is included in an attached file (Cover_letter_and_Response_to_reviewer_.docx). We kindly request that you open the document to view the responses.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Dear Authors,

Thanks for improvements. 

Your paper is accepted in the present version from my side.

Best regards

 

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.