Predictive Model of Granular Fertilizer Spreading Deposition Distribution Based on GA-GRNN Neural Network

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Comments and Suggestions for Authors

In this paper, the authors designed an UAV fertilizer particle deposition prediction system based on neural network decision-making, and the result provides a theoretical basis for precision fertilizer application operations by agricultural drones to a certain extent. Although the author had considered some of the factors in the research, additional explanations and descriptions of other key factors need to be added, as well as some formatting and grammatical issues in the article that need to be revised. The suggestions are as follows.

 

1.       L113, there is a duplication of the project number in the funding information. Similar mistakes appeared many times in the manuscript, please carefully verify and correct them one by one.

2.       Content of the TABLES: lack of standardization of letters. It is more reasonable that the first column “test number” in Table 1 changed to “Level”. The number of valid digits in Table 5 are not uniform. Please check all the Tables.

3.       Figure 1 showed that both the base frame and the spray system of UAV have an impact on spreader. Similarly, the test platform in Figure 5 also had scaffolds that may affect the dispersion of working fertilizer particles in spreader, but these factors were not considered in the prediction model, please add corresponding explanations and discussions.

4.       According to the author, the Centrifugal disk speed, working height, opening angle and operating speed were analyzed as the influencing factors. However, the influence of the rotor airflow on the dispersion effect of the powder particles during the UAV fertilization was ignored. Its rotor wind field is closely related to the flight parameters of the UAV, and the rotor airflow has a significant influence on the motion and dispersion characteristics of the classification. For this reason, the utility of the research content needs to be further verified.

5.       About the computation domain: L116 mentioned 50m * 50m * 10mm, obviously 10mm was wrong and needs to be corrected. Also, the superposition was determined to be 60m was shown in L150. In addition, the result part (L454) showed an effective width of only 7m. So, why was the length of the computation domain 50m selected?

6.       Figure 6 showed that the description is inconsistent with the original description, especially the placement spacing of the samples and the number of the samples, which were different from the original description.

7.       The conclusion part was not concise enough to highlight the focus of research.

Comments on the Quality of English Language

The quality of English needs improving.

Author Response

Dear reviewer,

Thank you for your comments on this article. I have made the necessary revisions, please refer to the attachment.

Best wishes.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The article addresses a topic of great interest in the current times in the field of applied agronomy, where UAVs are being widely used these days.

Several writing and methodological observations have been made that are relevant and can be found in the attached file of the manuscript, and that were indicated as highlighted comments in the text.

Most of them are crucial, but among others, a key observation is the following, and is located on page 12, line 377:

“Here is a conceptual error, or a poor explanation of the methodology in the previous paragraphs. The four factors are what the authors call the four operating parameters. Each of these factors has 4 levels (according to table 1 on page 3), so if all the combinations are compared there should be 256 tests (4×4×4×4). The authors set up 16 tests, for the 4 factors. In this case then the number of levels per factor should be 2. Table 3 is titled "Factor table for controlled trials", and indicates precisely 2 values ​​for each factor, but they are not the ones shown in the combinations of the graphs in figure 11 for each model evaluated. So what criteria were used to select these 16 tests from a possible 256 through a grid-search evaluation mechanism? This is not clear from the description of the methodology”.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer:
Thank you for your comments on this paper, for which I have made the appropriate changes. Please see the attachment.
Best wishes！

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have improved the manuscript sufficiently. There are only two minor observations to be taken into account. Both are detailed in the attached file as comments.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer:
Thank you for your comments and I have revised the paper accordingly. Please see the attachment.
Best wishes!

 

Author Response File: Author Response.docx