Fuzzy Neural Network Dynamic Inverse Control Strategy for Quadrotor UAV Based on Atmospheric Turbulence
Round 1
Reviewer 1 Report
The manuscript entitled "Fuzzy neural network dynamic inverse control strategy for quadrotor UAV based on atmospheric turbulence" has been investigated in detail. The topic addressed in the manuscript is potentially interesting and the manuscript contains some practical meanings, however, there are some issues which should be addressed by the authors:
ü The authors need to emphasize their contributions/novelties in the revision. In the current version, the authors did not discuss their contributions in detail.
ü The authors should carefully proofread this paper and correct all the typos in the revision. In the current version, there are still some typos/grammar errors.
ü Could the authors report the running time of the proposed algorithm? In this way, we can justify whether this algorithm can be applied to large-scale dataset.
ü The authors should address the following limitations. The first limitation is that, the related works should be grouped into two or three subsections. In the current version, the authors all merged them together.
Author Response
Response 1: Most of the past researches on resistance disturbances of quadrotor UAV were based on the gust, which is a discrete or definite wind speed change. In this paper, the atmospheric turbulence sequence is established by the Dryden model, and the continuous random wind is generated by the shaping filter, which can more truly simulate the air disturbance. As far as we know, this is the first time to use the dynamic inversion method of a fuzzy neural network to deal with the influence of atmospheric turbulence on quadrotor UAVs. And remarkable results were achieved.
Response 2: Thank you very much for your kind reminding. To get rid of grammar errors and typos, we have checked the whole paper carefully, and we believe the quality of the paper is greatly improved.
Response 3: The environment of the simulation experiments in this paper is a Core i7-9700F CPU, Windows 10 operating system, and a eight-core AMD processor. The physical memory is 16 GB, and the speed of the processor is 3.20 GHz. The code runs for 20 seconds. For the aircraft, if it can remain stable in the take-off stage, then the whole flight process can also ensure good flying quality. Therefore, the running time of 20 seconds can verify the performance of the controller.
Response 4: Thank you very much for your advice. According to the relevance of the content, the related work has been divided into corresponding subsections, which greatly improves the hierarchy of the article.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In the article under review, the authors presented the results of a study of the developed fuzzy neural network dynamic inverse controller for quadrotor unmanned aerial vehicles (UAV), which makes it possible to increase the stability of the position of the quadrotor UAV under conditions of atmospheric turbulence disturbances. The studies were carried out by the method of mathematical modeling.
In the Introduction and literature review, the prerequisites for conducting research are considered in sufficient detail, and the purpose of the paper is formulated. In the main parts of the paper, a mathematical description of the atmospheric wind field model, and a description of the dynamic model of a quadrocopter are proposed. The fuzzy neural network dynamic inverse controller has been developed. The simulation results are presented.
The results of the presented studies can be useful to a wide range of readers – scientists and practitioners, specialists in the field of control of quadcopter UAVs.
However, during the review, I was left with a few questions and I drew attention to the following shortcomings, the correction of which would improve the quality of the article:
- The paper repeatedly states that when using the PD-DIC method, there are always tracking errors that exist in the roll, pitch and yaw channels. However, the paper does not clearly explain the reason for this. I think these clarifications should be added.
- The authors do not present the results of any experimental verification of their studies. Please explain what the confidence in the adequacy of the obtained results of theoretical studies is based on.
- I invite the authors to additionally present in the form of a table the simulation results given in Section 5.3. This will allow the reader to perceive the material more clearly.
- What software did the authors use? I think it would be worth showing it.
It seems to me that the general shortcoming of the work is the lack of a description of the practical application of the results obtained. Nevertheless, I recommend the paper for acceptance after minor revisions.
Author Response
1. PD-DIC seeks to make the controlled object be intensively adjusted in the shortest time, while there will be adjustment static error. It can't perform adaptive control perfectly when there is external disturbance, and the anti-interference ability to the outside world is weak, so the tracking error always exists in the attitude channel.
2. The flight environment of the quadrotor UAV set in this paper is complex, and our experimental team does not have the corresponding equipment conditions for the time being, so the research methods mentioned and the experimental results obtained are based on the simulation platform. But we still guarantee the accuracy of the results through repeated rigorous verification. After the laboratory facilities are ripe, the corresponding experiments will be carried out.
3. Thank you very much for your kind reminding. After the simulation results in Section 5.3 are shown in tabular form, the experimental results are more intuitive and convincing.
4. MATLAB is not only an intuitive and efficient computer language, but also a scientific computing platform. It provides the core mathematical and advanced graphic tools for data analysis and visualization, algorithm and application development. According to more than 500 mathematical and engineering functions provided by it, engineers and scientists can interact or program in its integrated environment to complete their own calculations. When it comes to MATLAB, we can't help but mention Simulink. Simulink is a software used to model, simulate and analyze the dynamic system in the real world. Simulink provides a block diagram interface based on the data, graphics and programming functions of MATLAB core. Through block-to-block connection and attribute setting, users can easily build a model that meets specific requirements, and analyze and simulate the model.
Author Response File: Author Response.pdf
