A Deep Learning-Based Fault Warning Model for Exhaust Temperature Prediction and Fault Warning of Marine Diesel Engine

Round 1

Reviewer 1 Report

Firstly, the authors must be congratulated for undertaking an interesting and useful research project on hybrid prediction modeling for fault warning in marine Diesel engines considering the significant performance monitoring parameter, EGT. Following comments, suggestions are made to help authors improve the presentation quality and correcting few mistakes

1.      Line 460 -  There is no Table 9 found in the text?  Where it came from

2.      Line 530 -  Chapter 3 ? where it came from?

3.      Page 10, Fig. 6 – what is this Chineese title , Need to be removed?

4.      Line 287-288 , if each data point collected at 6 sec. interval and a total of 7200 data collected and used in the prediction model experimentation, then I guess total duration data were collected should be 12 hrs. Why it is mentioned 2 hrs.  Please check to avoid confusion.

5.      Section 4 Focuses on Experimentation with fault warning data. However, no simulation analysis results and discussion presented with the effectiveness of several prediction models used with the simulated faulty data, Why. If not then what is the need to include Sec. 4.

Few suggestions are included for future reseach

1.      Research should be carried out for another monitioring parameter, say for example Turbocharger EGT

2.      Prediction capability of the selected prediction models and methodologies with long term time-series data (say data collected over a month or so)

none

Author Response

Thank you very much for your valuable comments on this article. I have revised the article accordingly based on your suggestions. I have corrected all the suggestions and compiled the detailed changes in the attachment.Please see the attachment.Thanks again for your review comments.

Author Response File: Author Response.pdf

Reviewer 2 Report

A BiLSTM approach is presented for the temperature and fault prediction for marine diesel engines. I have some suggestions and comments that nee to be addressed. 

(1) The abstract need to be revised for the reflection of the need of the study and technical contribution of the proposed work. It is missing in the current version of the manuscript. 

(2) In the introduction section, there should be a space between text and giving reference i.e., maintenance [3]. Check this mistake all in the paper. 

(3) Line 46, try to abbreviate deep learning as (DL). Then use it in the paper. This mistake is done for other abbreviations. Need to see this carefully. 

(4) Line 57, it is suggested to use just a reference at the end. Avoid et. al. Line 58, what is CGMP? Abbreviations should be defined first. 

(5) Line 92, Cheliotis et al. combined the expected behavior(EB). Typos error, there should be space between behavior (EB). This mistake is all in the error. This needs to be checked. 

(6) Line 109 to 113, from Then...... Lengthy sentences and hard to follow. Make it clear and concise.  Line 155 and 156. Typos error, spacing issue. The why Padding? why not padding? Line 158, why not CNN.? Repetition of this kind of mistake is all in the paper.

(7) Line 174 to 178. Re-write this sentence where each unit operation is defined. Line 186, should be a capital Ot. 

(8) missing references such as should be inserted in the literature. Line 46 to 47. ''Han, Seung Heon, Tariq Rahim, and Soo Young Shin. "Detection of faults in solar panels using deep learning." 2021 International Conference on Electronics, Information, and Communication (ICEIC). IEEE, 2021.     ''Shao, Siyu, et al. "Highly accurate machine fault diagnosis using deep transfer learning." IEEE Transactions on Industrial Informatics 15.4 (2018): 2446-2455.

(9) Lines 190 to 193, are hard to follow. Re-write. Line 239 to 240, too many and and and......

(10) Figure 6, remove the Chinese legends on top. Figure 8 is an extremely poor representation of the results as to which color is applied for which approach. It is hard to differentiate. 

(11) 

 

 

The paper is full of grammatical of expressions mistakes and requires moderate English proofreading. 

Author Response

Thank you very much for your valuable comments on this article. I have revised the article accordingly based on your suggestions. I have corrected all the suggestions and compiled the detailed changes in the attachment.Please see the attachment.Thanks again for your review comments.

Author Response File: Author Response.pdf