Review Reports

Round 1

Reviewer 1 Report

Please refer to the detailed comments file.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors describe a control algorithm for a pH-control of a liquid fertilizer which is  a common combination of  irrigation and fertilization in agriculture. The specific control problem is in the relatively long delay time within the control loop which is manifested by a fuzzy-PID-Smith Predictive compensation algorithm.

Remarks:

1. In the introduction the authors review mostly the literature on special fuzzy control cases regarding PID fuzzy controllers and smith predictors in process control. However,  tuning and optimization of fuzzy controllers especially in hierarchical  process control structures should be mentioned. See here   

B. Demaya, R. Palm, S. Boverie,  A. Titli: Multilevel qualitative and numerical optimization of fuzzy controllers. Proc. of the 1995 Intern. Conf. on Fuzzy Systems, Fuzz IEEE 1995.

T. Nalovsky: Optimization of Fuzzy Controller Parameters for the Temperature Control of Superheated Steam. Procedia Engineering, Vol 100, 2015, pp.1547-1555  

2. Page 4: In the description of the static model a corresponding reference should be given because we have the problem that control engineers may not be familiar with chemical reactions and formulas. So it would be helpful to explain some more steps a little better especially the balance equations.

3. Page 6, equation (13): Write here the general form like                           G(s)= K*exp(-τ*s)/(1+sT)  and then refer to the variables according to their numbers

4. Page 7: Refer to some reference regarding to the Cohen Coon Method (from 1953). Many control engineers may work with the Ziegler/Nichols Method (from 1942) which is a heuristic method similar but inferior to Cohen Coon. Mention this in  this chapter.

5. Page 7: Explain Table 1 better: for example                                      "PM/NB/NB regards to Kp/Ti/Td"

6. Page 9: In Fig 7 a * is missing:  G_0^*(s) in the feedback loop.

7. Page 10:   In the "denominaters of the  denominators" eqs. (19) and (20) the exp(-τ*s) must be cancelled. See also eq. (22) where this term does not exist.

8. Page 10: Give a correspondence between Figs. 7 and 8 showing which parts of Fig. 7 correspond to which part in the simulation model in Fig. 8.

The problem I see is that if you make G_C2 too small, the connection between Y and G_C2 is cut off. Then we have the strange situation, where plus and minus "compete" at the node at the output, with minus "wins". This could in turn have an impact on the feedback which becomes positive. 

In addition, the system dynamics is determined by the model G_0^*. So, in the simulation this might work, but what about the experiment itself?

 

9. Page 11: In table 1 only the controller is described. Where and how (in which control scheme) are the performance indicators explicitely described?

10: Page 17-18: What does the [J] at the end of a paper title mean?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The quality of the manuscript is improved, and the revised version of the manuscript is recommended for publication.