A Segmented Adaptive PID Temperature Control Method Suitable for Industrial Dispensing System

Round 1

Reviewer 1 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

 

First of all, I would like to thank the authors for their efforts in improving the manuscript. However, in my opinion, significant revisions are still required. The manuscript continues to exhibit a considerable number of grammatical issues and typographical inconsistencies, which reflect a lack of attention and detract from the clarity of the work.

Regarding the process under control, it is essential to include a dedicated figure that clearly illustrates the process diagram. This would allow the reader to better understand the interrelation of variables and the physical context of the system being analyzed. The current figures are primarily control-oriented and do not provide a descriptive representation of the actual process. Furthermore, it is imperative to include a photograph or schematic of the real experimental setup, rather than using a generic or conceptual image.

There are also conceptual issues that must be addressed. For example, the authors claim that nonlinear models incorporating temporal dynamics, when linearized, can result in high-order systems. This assertion is incorrect. Linearization does not increase system order; the number of state variables remains the same, as supported by standard nonlinear system theory.

In addition, the manuscript lacks consistency in the definition and usage of acronyms—many are redefined multiple times, which reduces the overall formality and coherence of the document. There are also persistent inconsistencies between the equations and their corresponding references in the text. For instance, equation (4) includes the product term "KA", while the text refers to it as "K*A", which may confuse the reader. Moreover, while variables and constants are correctly italicized within equations, the same formatting is not preserved in the main text, which reflects a lack of attention to detail.

From a modeling standpoint, the authors introduce the FOPDT model as a central element of their analysis but fail to justify its selection. A rigorous explanation for its application in this context is necessary to support its validity.

Figure 3 lacks proper labeling of axis units, which is a fundamental requirement for clarity. Likewise, the flowcharts presented in Figures 4, 5, 7, and 9 are difficult to interpret. The nature and type of signals represented by the arrows are unclear, suggesting that different types of variables (e.g., physical signals vs. control variables) may be improperly mixed.

In conclusion, the manuscript requires a thorough revision addressing the issues highlighted above before it can be considered for publication.

Author Response

Dear Reviewer, Greetings! Please find attached the point-by-point response to your comments in the attachment. We sincerely appreciate your valuable feedback and have carefully addressed each suggestion to improve the manuscript. Thank you for your kind consideration.

Author Response File: Author Response.pdf

Reviewer 2 Report (Previous Reviewer 4)

Comments and Suggestions for Authors

Check amplitude "A" in lines 135/137/139. Is that vector matrix given, or is it just parameter (function) and it should be written in italic (please just unify it in the whole text)...

Check writing of time stamps at lines 149 & 150, use index, and write ms in non-italic (straight)

Line 279 write italic (k+1) at the end

Line 428 --> 5.4 Disturbance

Check math type font size through text.

Reference list could be longer.

Author Response

Dear Reviewer, Greetings! Please find attached the point-by-point response to your comments in the attachment. We sincerely appreciate your valuable feedback and have carefully addressed each suggestion to improve the manuscript. Thank you for your kind consideration.

Author Response File: Author Response.pdf

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Please check the attached file.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer, Greetings! Please find attached the point-by-point response to your comments in the attachment. We sincerely appreciate your valuable feedback and have carefully addressed each suggestion to improve the manuscript. Thank you for your kind consideration.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

Dear Authors,

Thank you for the effort shown to correct this version. There are some minor details to be corrected. Particularly of form and not of substance. Re-read the version and you will see some details in the numbering of the equations.

Author Response

1. Summary
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions highlighted changes in the re-submitted files.
3. Point-by-point response to Comments and Suggestions for Authors
Comments 1: Thank you for the effort shown to correct this version. There are some minor details to be corrected. Particularly of form and not of substance. Re-read the version and you will see some details in the numbering of the equations.
Response 1: Thank you for pointing this out. We agree with your point of view. Upon careful inspection, it was found that the font size and alignment of individual formula serial numbers were not fully standardized. Therefore, all formula numbers have been thoroughly checked and revised in accordance with the format samples provided by the publisher.

Author Response File: Author Response.docx

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Although the authors added the fuzzy PID algorithm in the performance comparison, the reviewer still does not see the quantified performance analysis (e.g. ISE and IAE or time domain performance index).

For the academic community, all published paper results should serve as benchmarks for the performance comparison of subsequent papers. Therefore, only the overshoot index is not enough. The authors have to demonstrate the advantage of the proposed method with a detailed quantified performance analysis.

Author Response

1. Summary
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions highlighted changes in the re-submitted files.
2. Point-by-point response to Comments and Suggestions for Authors
Comments 1: Although the authors added the fuzzy PID algorithm in the performance comparison, the reviewer still does not see the quantified performance analysis (e.g. ISE and IAE or time domain performance index). For the academic community, all published paper results should serve as benchmarks for the performance comparison of subsequent papers. Therefore, only the overshoot index is not enough. The authors have to demonstrate the advantage of the proposed method with a detailed quantified performance analysis.
Response 1: Thank you for pointing this out. We agree with this comment and have added Section 5.5 to the article to calculate the changes in ISE indices of the three control methods at different stages. Additionally, we have included an analysis of the experimental results for this section in Section 5.6. You can find these revisions in lines 415 to 425 and lines 455 to 472.

Author Response File: Author Response.pdf

Round 3

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

No further comment.

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.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I thank the authors for writing up the results of the research and the editor for giving me the opportunity to review the manuscript.

The manuscript presents a study of temperature control of an industrial glue dispenser. The envisioned control algorithm is to be computationally efficient, reach the target temperature fast and without overshoot, and hold the desired temperature well. The goals are achieved with a PID, the parameters of which change depending on the current temperature, and the integral term is carefully initialised. The setup is experimentally shown to perform well.

The study successfully addresses a real problem experimentally, making it valuable.

However, there are significant omissions in the presentation of the study that seriously limit its usability. While the Introduction is relatively good in this aspect, line 51 is deceiving: it implies that the references [12],[13],[14] are about adhesive dispensing, but they are not. Please fix it. We do learn the purpose of your study though: "This method can reduce the overshoot and accelerate the convergence speed of the algorithm while maintaining a relatively low computational load."

The experimental setup is shown in Figure 1 but never sufficiently described. I believe "DTU" should be "DUT". The whole study describes a system with a single input and single output, while in Figure 1 we see 5 sensors and 5 actuators – how come? It cannot be seen in Figure 1, but the text explains how the glue travels from the pressure plate to the glue gun valve – I think it should be illustrated. Which heating plate and which thermocouple does your PID use? I can't find it in the text. Are you doing the measurements under the industrially relevant conditions with the glue flowing? Is the glue flow a disturbance of the system, or can it be neglected for some reason (such as that the flow and the temperature of incoming glue are constant)? All this information is very important and should be included.

What are the two lines in Figure 2? Is the solid line the actual measurement and the dashed line is the result of the simulation of a model with the parameters fitted to the observations? If so, how do you obtain the model parameters?

Is the data in Figure 3 coming from measurements or from theory? What about 5 (a) and 6, 8, 10, 11, 12? There must be no doubt, it has to be specified directly.

Reference 26 is never cited.

Lines 194–198 repeat what was just said in the few lines before.

Section numbering is inconsistent. There are two sections 3, and there is 2.3 in line 216 in a section 3.

What does the word "proprietary" in line 346 mean, is it just the wrong word?

Is "below 10°C" in line 252 supposed to be "within 10°C" or is it really this cold? Are you sure that negative 100°C is meant in line 353?

What do you mean by "the traditional PID algorithm limits the maximum value of the integral value"? Is there such a limit in traditional PID algorithms, and is 80 a standard for it, or is this the best limit you found for your example?

There is no proper discussion of the results. Discussion is a crucial part of every successful research article. It should explain what do the results mean and how do they advance the field. The discussion on how does the research fit in the context of existing literature (that is properly cited) tells the reader if and why to consider your article in their work. Please write it.

The Conclusions section could be used better. It is supposed to tell the reader the main points that can be learned from the work you have done. It should therefore be based on your results. However, the first part of the Conclusions, in lines 403–409, is instead wasted on reiterating what was already known before your work was done. In line 413, we learn that the algorithm "has simple calculations" – can you quantify it? I know it was your goal to keep the method computationally simple, so one would expect an explanation of how successful you were among your results (and in the discussion), but I can't find it. The part in the lines 416–419 is repeated from the Abstract, which is also wasteful as it tells nothing new to the reader that has presumably read the Abstract already. It is also hard to interpret for someone that hasn't read the rest. The absolute numbers of 2 to 4℃ overshoot don't tell much without knowing what the temperature change is, neither is the 0.2℃ fluctuation without quantifying the disturbance. The definition of the 30% increase in convergence speed is also unclear, and I can't figure out what result it refers to. Maybe you can elaborate more on these important conclusions.

It is very nice that you make the data available. However, there is no metadata (at least not in English), we don't know what the numbers mean. Can you document how they were collected, please? Also, the Instructions for Authors at https://www.mdpi.com/journal/electronics/instructions#suppmaterials require that "For work where novel computer code was developed, authors should release the code either by depositing in a recognized, public repository or uploading as supplementary information to the publication." Can you provide the controller and the data-processing code you wrote?

The manuscript is also seriously lacking in the field of theory, which must be fixed. The findings that a PID cannot get to the operating point swiftly without an overshoot and at the same time keep the temperature stable in spite of the disturbances, and the improvement through the adjustment of PID parameters, are presented as original ideas contributed by this work. In reality, they are trivial consequences of control theory you seem not to be familiar with, also judging by the terminology you use. It is a widely-known fact in control that there is a trade-off between the controller's ability to track the setpoint and to reject disturbances. A good setpoint-tracking controller is by necessity not a good disturbance-rejection controller, and vice versa. For you, setpoint tracking is important at startup, when you want to reach the new reference temperature swiftly and without too much of an overshoot. During the operation, however, you care about disturbance rejection (you seem to invent the term "Anti-interference performance" for disturbance rejection). As it is impossible to do both with the same control law, you have to re-tune your PID between both system states. This is also a standard approach, known as gain scheduling. Therefore, the ideas you present as original are actually a direct application of control theory to your particular problem, while you make it look like you are not familiar with control theory. This does not look good and is not beneficial to the reader. Please provide the background on control theory in the introduction, explain how your methods fit into the theory, and use standard terminology (disturbance rejection in place of anti-interference performance, call "setpoint tracking" and "gain scheduling" their proper names).

Another detail I noticed is that you mention robustness in the introduction but not elsewhere in the text. If robustness is not a topic of your work, leave it out from the Introduction; if it is, elaborate on it in the text. In case that you are not aware, there tends to be a trade-off between robustness and performance (similarly than between setpoint tracking and disturbance rejection): the best-performing controller is not robust.

I hope the review helps you with improving the manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

In my opinion, the manuscript requires substantial improvement in both technical content and formal presentation.

Substantive Issues:

• The abstract lacks clarity and is overly convoluted, making it difficult to follow the main contributions of the work.

• A paragraph outlining the manuscript structure is missing, which hinders the logical flow.

• All system equations should be presented explicitly in the time domain, especially given their likely nonlinear nature. Additionally, complete transfer functions that represent the system dynamics must be included.

• There is a lack of consistency between equations in the Laplace (continuous) domain and those in the discrete (k-indexed) domain, leading to a breakdown in the narrative and conceptual flow.

• The tuning methodologies for the control algorithms are not described and should be clearly presented.

• The concept of "system output power" needs to be formally defined and physically explained.

Formal and Stylistic Issues:

• Acronyms should be defined only once, and only if they are consistently used throughout the manuscript.

• Many paragraphs contain spelling and grammar errors, with repetitive and disorganized phrasing that detracts from readability.

• There are formatting issues at the beginning of paragraphs, including inconsistent indentation and typographical errors.

• Figures require significant enhancement: axes must include units, and the formatting—especially of the x-axis—needs refinement.

• Equation numbering is incorrect, and equation formatting is inconsistent, including inconsistent use of the multiplication operator and variable styles (italicized vs. plain).

• All existing text should undergo thorough revision for clarity, coherence, and conciseness.

• In several cases, random capitalization appears mid-sentence without punctuation; these errors must be corrected.

• Figures should be introduced in the text before being displayed, following standard academic writing conventions.

• The manuscript should be written in the third person, avoiding first-person expressions unless strictly justified.

• The origin of the constants used in (7) should be described and justified based on physical or experimental data.

• Section headings (e.g., Section 3) should avoid double punctuation marks like “:”.

• The control flow diagrams have issues with decision symbols and should be reviewed for correctness.

• The nature of the PID actions (direct vs. indirect control) must be specified.

• All current values should be expressed with their respective units (i.e., amperes [A]).

• A photograph of the experimental setup should be included to support the practical implementation claims.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors propose a segmented adaptive PID temperature control method for industrial dispensing system. Although the proposed method seems interesting, the transfer function of plant is too ordinary to evaluate the performance of proposed method.

On the other hand, there is no quantified performance analysis (e.g. ISE and IAE or time domain performance index) to demonstrate the advantage of proposed method. Besides, there is not enough that just only traditional PID be compared. The authors have to compare more algorithms in performance comparison.

The following issues should be modified:

1. There is an inconsistency in the decision labels “YES” and “NO” in the flowchart of Figure 4. In addition, Figure 7 seems to include an extra “NO” label. The authors have to revise the flowcharts.
2. The axis labels, units, and error bounds in Figures 10 to 12 are either missing or incomplete. For clarity and readability, the authors should ensure that all figures include comprehensive and standardized annotations.
3. Section 4 lacks a detailed discussion of the parameter tuning process for both the traditional PID and the proposed method. The values or selection criteria for Ki, Kp, Kd, α, and β are not provided. Clarifying how to tune the parameters can enhance the transparency and comparability of the experimental results.

Reviewer 4 Report

Comments and Suggestions for Authors

The paper is OK. I have just comments on visibility of lines in diagram if someone wants to print paper to study it. 

You can find comments in the attached document.

Comments for author File: Comments.pdf