Evaluation of Artificial Neural Network to Model Performance Attributes of a Mechanization Unit (Tractor-Chisel Plow) under Different Working Variables
Round 1
Reviewer 1 Report
The article considers the use of artificial neural network model and multiple linear regression methods to model the characteristics of the mechanization unit (tractor-chisel plow) during plowing process, taking into account working variables and different soil types. The proposed method gives thorough results.
However, there are some comments. And authors should take into account this comments to correct their article.
- It is not desirable to write the obtained numerical results of the characteristics in the annotation, instead it is better to indicate it in the conclusions. The annotation should summarize the results in terms of the character of changes of the obtained coefficients values​.
- Given the amount of used literature, it is necessary to change the structure of the article, in particular to add a section "Relation works"
- In my opinion, it it is necessary to depict the structure of the neural network layers. It is necessary to justify the choice of the number of layers more precisely.
- It is desirable to consider in more detail the future of the work, in particular to indicate the prospects for the application of research and results.
Author Response
We greatly respect all your comments.
Thanks!
Author Response File: Author Response.pdf
Reviewer 2 Report
The results presented in the article under review are of interest to researchers of the processes of tillage by agricultural machines. Such results contribute to the reduction of energy costs in agriculture. Therefore, the topic under study seems to be relevant and justified.
At the same time, the proposed method for processing statistical data based on ANN is debatable. The purpose and scientific novelty of the work also require clarification.
There are a number of questions to the presented analysis of the results:
- in fig. 1, the vast majority of the analyzed points are in the region of the soil index 0 - 20. With such a non-monotonic distribution of points, linearization is a bad solution,
- can FWI be > 120?,
- distribution in fig. 2 is also far from linear,
- how did the authors determine during the experiments the engine power and other parameters related to the determination of energy consumption for field work?
Author Response
We greatly respect all your comments.
Thanks!
Author Response File: Author Response.pdf