Chaotic Analysis and Prediction of Wind Speed Based on Wavelet Decomposition

Round 1

Reviewer 1 Report

Dear Authors,

It was very interesting to review such a thorough mathematical and technical study. Undoubtedly, the results obtained are of great theoretical and applied importance in the field of modeling and research of wind flows. However, there are a number of reasons why I cannot recommend the article for publication in its current form.

1. In the abstract on line 18, the point is duplicated.

2. In fig. 3 and 4, the x-axis can be designated as time, indicating the units (days, months).

3.. As I understand it, Figure 3 is a generalization of Figure 4. If yes, then both graphs can be plotted in the same axes in the same figure.

4. How was the initial difference between the points chosen in the LLE calculation algorithm?

5. In fig. 12 part of the text is lost.

6. Is the emergence of multistability possible in the proposed model? Since in this case, to determine the presence of chaos in the oscillations, it is worth using other tools, for example, relative specific volume from recurrence analysis. Please compare the simulation results by implementing this metric instead of LLE as a way to detect chaos. 

7. In the introduction, it is said that when modeling winds, it is important to investigate long time intervals. It is required to evaluate how the proposed model behaves in long-term simulation.

8. English improvement required.

Author Response

see attached.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper  studies  characteristic of wind speed  in direction  to use it speed prediction of the wind speed.. Various methods are combined and it  is not clear whether this is reasonable. All used methds works under some assumptions- this is not done, arguments go through the data. The application to the data looks reasonable.

 Comments:

P.8, l.211-- what is w_k?

Figure  12 -- how is made decision in  "yes" or "no" "chaotic?

Author Response

please see attached.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear Authors,
You have done a great job revising the manuscript. Thank you for taking my recommendations into account.

  
The LLE method to detect chaotic oscillations was asked in the context of the possible multistable behavior of the model. When calculating the LLE, close initial conditions are used, and in the case of multistable systems, trajectories from such values can lead to completely different attractors, including periodic ones. This can be confusing results. Therefore, this approach really should not be used as the only tool to study chaotic systems. The relative specific volume metric was considered in the paper "Detection of Hidden Oscillations in Systems Without Equilibrium" IJBC, 31.03 (2021): 2150043. In this study, it is shown that the method for LLE calculation can give erroneous results. Therefore, it would be interesting to compare LLE results and the relative specific volume to study the proposed model.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

No futher comments from my side

Author Response

The authors sincerely appreciate the effort by the reviewer.