An Unmanned Aerial Vehicle Troubleshooting Mode Selection Method Based on SIF-SVM with Fault Phenomena Text Record
, Wenbing Chang 2 and Shenghan Zhou 2,*Round 1
Reviewer 1 Report
In this work is presented this a mode selecting method based on SIF‐SVM with the fault phenomenon description text data.
The following issues must be attender in order to improve the quality of the manuscript.
The novelty and contribution of the proposal have to be highlighted, also, it is necessary to emphasize the problems that have not been solved and that are addressed through this proposal.
It is recommended to perform the writing of the manuscript in third person instead of first person. (i.e. lines 276 and 281).
Figure 5 shows the modeling procedure and model application procedure, however, its step-by-step description must be improved.
The conclusion section has to be rewritten since its present form are not convincing, additionally, future works may be mentioned.
Author Response
Dear Reviewer,
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript. we appreciate you very much for the positive and constructive comments and suggestions on our manuscript entitled “A UAV Troubleshooting Mode Selecting Method Based on SIF-SVM with Fault Phenomena Text Record”. (Manuscript ID: aerospace-1400257).
We have studied reviewer’s comments carefully and have made revision in the paper. The main corrections in the paper and the responds to the reviewer’s comments are as following:
Point 1: The novelty and contribution of the proposal have to be highlighted, also, it is necessary to emphasize the problems that have not been solved and that are addressed through this proposal.
Response 1: Thank you for your comments. At present, if the UAV fails, professional technicians and equipment are required to locate the fault and provide maintenance scheme. However, in the UAV flight site, there is often a lack of professional technicians and equipment. The method we proposed is to locate the fault system and select the treatment mode according to the actual process and the description of the fault by the field personnel, without professional inspection. At present, there is no research and practice in this field. This method is of great significance for fault diagnosis and maintenance in UAV mission site. It can improve maintenance efficiency, reduce maintenance cost and ensure flight safety. We also emphasized this point again in the introduction.
Point 2: It is recommended to perform the writing of the manuscript in third person instead of first person. (i.e. lines 276 and 281).
Response 2: Thank you for your comments. This suggestion is very necessary. We have made corresponding modifications in the paper.
Point 3: Figure 5 shows the modeling procedure and model application procedure, however, its step-by-step description must be improved.
Response 3: Thank you for your comments. Based on your suggestions, we explained the modeling process and added the following contents: “Based on the UAV fault record text data, the UAV fault location model and troubleshooting mode selection model were established. In the modeling, the text description of fault phenomenon was preprocessed, and feature extraction was carried out. Then, using the text feature data and fault location label, the fault location model was established via the SVM. Next, the serial information fusion method was used to fuse the fault phenomenon text feature data with the fault location labels, and, thus, the troubleshooting mode selection model was established by using the troubleshooting mode labels and the SVM. In the application of the model, through the fault location model, the fault location was determined by using the text feature data from the fault phenomenon. Once the fault location and fault phenomenon text were fused, the fused data was used to select the troubleshooting mode through the troubleshooting mode selection model.”
Point 4: The conclusion section has to be rewritten since its present form are not convincing, additionally, future works may be mentioned.
Response 4: Thank you for your comments. According to your opinion, we have revised the conclusion. In the future work, the method proposed in this paper can be applied to the practice of a certain type of UAV. Moreover, according to a large number of actual cases, the trouble-shooting modes can be divided more finely to realize more detailed guidance to on-site personnel.
Sincerely yours,
Linchao Yang on behalf of the authors.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This paper is well written. There are some comments from reviewer.
- The introduction part should divided into two parts: introduction and literature review.
- The literature review should discuss the recent method:
- Robust fault tolerant controller design for Takagi-Sugeno systems under input saturation. International Journal of Systems Science.
- Adaptive sliding mode control for attitude and altitude system of a quadcopter UAV via neural network. IEEE ACCESS.
- Finite-Time Attitude Fault Tolerant Control of Quadcopter via Neural Networks. Mathematics - What is the advantage compared to other method?
- The contribution should focus more.
- The References should follow the format of journal.
- Figure 5 is not clear.
- The conclusion should discuss the future work.
Author Response
Dear Reviewer,
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript. we appreciate you very much for the positive and constructive comments and suggestions on our manuscript entitled “A UAV Troubleshooting Mode Selecting Method Based on SIF-SVM with Fault Phenomena Text Record”. (Manuscript ID: aerospace-1400257).
We have studied reviewer’s comments carefully and have made revision in the paper. The main corrections in the paper and the responds to the reviewer’s comments are as following:
Point 1: The introduction part should divided into two parts: introduction and literature review.
Response 1: Thank you for your comments. This suggestion is very necessary. We have divided the introduction in two introduction and literature review and improved them.
Point 2: The literature review should discuss the recent method:
- Robust fault tolerant controller design for Takagi-Sugeno systems under input saturation. International Journal of Systems Science.
- Adaptive sliding mode control for attitude and altitude system of a quadcopter UAV via neural network. IEEE ACCESS.
- Finite-Time Attitude Fault Tolerant Control of Quadcopter via Neural Networks. Mathematics
Response 2: Thank you for your comments. We have carefully read the above papers and discussed them in this paper.
Point 3: What is the advantage compared to other method?
Response 3: Thank you for your comments. At present, if the UAV fails, professional technicians and equipment are required to locate the fault and provide maintenance scheme. However, in the UAV flight site, there is often a lack of professional technicians and equipment. The method we proposed is to locate the fault system and select the treatment mode according to the actual process and the description of the fault by the field personnel, without professional inspection. At present, there is no research and practice in this field.
Point 4: The contribution should focus more.
Response 4: This method is of great significance for fault diagnosis and maintenance in UAV mission site. It can improve maintenance efficiency, reduce maintenance cost and ensure flight safety.
Point 5: The References should follow the format of journal.
Response 5: Thank you for your comments. We have revised some of the non-standard expressions.
Point 6: Figure 5 is not clear.
Response 6: Thank you for your comments. We have modified Figure 5.
Point 7: The conclusion should discuss the future work.
Response 7: Thank you for your comments. In the future work, the method proposed in this paper can be applied to the practice of a certain type of UAV. Moreover, according to a large number of actual cases, the trouble-shooting modes can be divided more finely to realize more detailed guidance to on-site personnel.
Sincerely yours,
Linchao Yang on behalf of the authors.
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper proposes a mode selection method for the fault analysis of UAVs based on a serial information fusion support vector machine with fault phenomenon description text data.
In general, the approach is interesting but several issues prohibit publication in this form.
Firstly, the authors fail to explain one crucial point: why should a UAV send plain text? Usually, there are certain defined parameters that are exchanged such as sensor signals but also signals identifying the operation and health of certain components. Also detected faults should correspond to certain predefined tables and can be handled as parameters. Why choosing any other form of communication? How to combine any other forms of communication with a structured form of communication? The authors need to provide answers to these questions.
Secondly, the authors do not use to established terms for fault analysis such as fault identification, fault detection etc. They need to base their discussion on the given terminology of fault diagnosis (see e.g Blanke et al.).
Thirdly, the text is to abstract. What are the faults, the authors wish to analyze? They need to be presented best in a table. Which new fixed wing UAV is described? A reader needs to know the general design in order to be able to understand the possible faults. The new fixed wing UAV needs to be described and the design description needs to be linked to the potential faults.
It must be made clear that it would not be possible to assign a troubleshooting mode by means of e.g. a simple lockup-table including certain faults and corresponding troubleshooting modes.
A few smaller remarks:
In the title and the abstract I would recommend to avoid any abbreviations
Also in the text all abbreviations need to be explained
Author Response
Dear Reviewer,
On behalf of my co-authors, we thank you very much for giving us an opportunity to revise our manuscript. we appreciate you very much for the positive and constructive comments and suggestions on our manuscript entitled “A UAV Troubleshooting Mode Selecting Method Based on SIF-SVM with Fault Phenomena Text Record”. (Manuscript ID: aerospace-1400257).
We have studied reviewer’s comments carefully and have made revision in the paper. The main corrections in the paper and the responds to the reviewer’s comments are as following:
Point 1: Firstly, the authors fail to explain one crucial point: why should a UAV send plain text? Usually, there are certain defined parameters that are exchanged such as sensor signals but also signals identifying the operation and health of certain components. Also detected faults should correspond to certain predefined tables and can be handled as parameters. Why choosing any other form of communication? How to combine any other forms of communication with a structured form of communication? The authors need to provide answers to these questions.
Response 1: Thank you for your comments. The amount of original flight parameter record data is large, and special interpretation tools need to be used to analyze it. In particular, this analysis can only be conducted for specific UAVs, not for a fleet (efficiency, interpretation mode, comparability, etc. are not supported). Fault text data is recorded by ground personnel and is a direct description of the fault. Since no special airborne equipment is required for recording, the availability of text data is very high. On the other hand, the amount of flight data is very large, so the cost of computing resources for analyzing flight data is very high. In practice, a large number of flight parameter data will not be stored for a long time, while the text data will be saved for a long time.
Because text data belongs to unstructured data, there are some difficulties in the interaction with structured information, especially for the UAV fleet at the swarm level. At present, there is no good method to carry out this kind of research, which can be tried in the future work.
Point 2: Secondly, the authors do not use to established terms for fault analysis such as fault identification, fault detection etc. They need to base their discussion on the given terminology of fault diagnosis (see e.g Blanke et al.).
Response 2: Yes, we deliberately do not use these terms to avoid misleading readers. Our research is fault analysis through data-driven method, rather than fault diagnosis and identification through fault mechanism in the traditional sense. Our research is more like the analysis of fault law.
Point 3: Thirdly, the text is to abstract. What are the faults, the authors wish to analyze? They need to be presented best in a table. Which new fixed wing UAV is described? A reader needs to know the general design in order to be able to understand the possible faults. The new fixed wing UAV needs to be described and the design description needs to be linked to the potential faults.
Response 3: Thank you for your comments. If we are trying to understand specific failure patterns, we may need to analyze structured data such as flight parameters. However, our research is not to analyze the fault mechanism of a certain type of fault. We would like to analyze the fault law on a more macro scale, such as whether the fault phenomenon shown by the fault has a centralized law, or what kind of troubleshooting strategy should be adopted according to the law when the fault phenomenon occurs.
Point 4: It must be made clear that it would not be possible to assign a troubleshooting mode by means of e.g. a simple lockup-table including certain faults and corresponding troubleshooting modes.
Response 4: Detailed and specific troubleshooting methods need to be judged by professional equipment and personnel. In the field of UAV use, there is often a lack of professional equipment and technicians. Our research would like to get the basic troubleshooting modes through the description of fault phenomena by field personnel. In this way, in practice, for some faults classified as on-site treatment, on-site personnel could not carry out complex and professional inspection, which would improve the use efficiency of UAV and reduce the inspection cost. Therefore, we used this method to classify the troubleshooting methods according to the expert experience and the requirements of UAV use site, which could meet the classification requirements of UAV in the practice.
Point 5: In the title and the abstract I would recommend to avoid any abbreviations. Also in the text all abbreviations need to be explained.
Response 5: Thank you for your comments. We have revised the title of the article and explained the abbreviations in the article.
Sincerely yours,
Linchao Yang on behalf of the authors.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The author handled all reviewer's comment. Therefore, this paper can be accepted for publication.
Reviewer 3 Report
Thanks for the responses. I think the responses are sensible. However, they need to be sensibly integrated in the text, because other cautious readers may face exactly the same questions as mine. I also did not see a title without abbreviations.
