Fault Diagnosis of Diesel Engine Valve Clearance Based on Variational Mode Decomposition and Random Forest
, Donghai Wei 3, Haipeng Zhao 2, Jinjie Zhang 1 and Zhinong Jiang 2Round 1
Reviewer 1 Report
Found mostly stilistic problems and mistypings. Please find attached document with markings.
Comments for author File: 
Comments.pdf
Author Response
Response to Reviewer Comments
Point 1: Found mostly stilistic problems and mistypings. Please find attached document with markings.
Response 1: Many thanks to the reviewer. We are sorry that we had made many typos in the previous paper. In the revised manuscript, we have made corrections and improved our English expression to the best of our ability.
We have replaced table 6 in the way you suggested. We've replaced table 6 with “Three different operating speeds (1500rpm, 1800rpm, 2100rpm) were investigated, and for every normal or fault conditions, there were 50 training and 20 test samples.”
According to your suggestion, we also added the schematic diagram of spring and other parts in figure 7.
Figure 7. Sketch of valve train.(Please see the attachment)
Author Response File: Author Response.pdf
Reviewer 2 Report
Dear Authors,
There are numerous places in the text with English grammatical errors. The authors should be full checking for grammar and mistakes to meet the quality of Journal.
Most equations and tables in the paper, the authors don't mention to any references. Add citation to the equations which took from references.
Add a new flowchart to explain the details of the proposed approach. There are many research papers study the same problem which investigated in the present paper. What is exactly the new point of this work?
The authors should focus to clarify this issue in the paper.
The meaning of the conclusions is unclear, and there are grammatical errors. The authors should think over the real significance of their results and try to rewrite this section to improve understanding of the conclusions.
Altogether, the paper needs modification to be suitable for the standards required for publication; therefore I recommend that it required to major revision.
Author Response
Response to Reviewer Comments
Point 1: There are numerous places in the text with English grammatical errors. The authors should be full checking for grammar and mistakes to meet the quality of Journal.
Response 1: Many thanks to the reviewer. We are sorry that we had made many typos in the previous paper. In the revised manuscript, we have made corrections and improved our English expression to the best of our ability.
Point 2: Most equations and tables in the paper, the authors don't mention to any references. Add citation to the equations which took from references.
Response 2: We thank the reviewer for the kind suggestions. Aiming at the deficiency of formula references, we add three references. References are as follows:
K. Dragomiretskiy et al. Variational mode decomposition. IEEE Transactions on Signal Processing. 2014, 62, 531-544. F. H. Li et al. Data-driven time-frequency analysis method based on variational mode decomposition and its application to gear fault diagnosis in variable working conditions. Mechanical Systems and Signal Processing. 2019, 116, 462-479. X. X. Jiang et al. A coarse-to-fine decomposing strategy of VMD for extraction of weak repetitive transients in fault diagnosis of rotating machines. Mechanical Systems and Signal Processing. 2019, 116, 668-692.
Point 3: Add a new flowchart to explain the details of the proposed approach. There are many research papers study the same problem which investigated in the present paper. What is exactly the new point of this work?
The authors should focus to clarify this issue in the paper.
Response 3: The issues that the reviewer mentioned are very important and the suggestions are very helpful to make substantial improvement on the paper. We have improved the flow chart to make it more detailed.
Figure 6. Fault diagnosis process.(Please see the attachment)
According to reviewer's opinions on innovation points, we added a note of innovation in the introduction section. In the revision, the corresponding modifications are as follows:
“The main contribution of this work can be summarized as follows: 1) Based on reducing the energy leakage of the IMF component frequency spectrum and the minimum IMF component of power spectral entropy, the VMD penalty function is optimized; 2) RF algorithm is adopted to rank the importance of all extracted features, so as to reduce dimensions and improve calculation efficiency; 3) The combined VMD and RF method improves the accuracy of fault identification.”
Point 4: The meaning of the conclusions is unclear, and there are grammatical errors. The authors should think over the real significance of their results and try to rewrite this section to improve understanding of the conclusions.
Response 4: We thank the reviewers for the valuable comments on the conclusion of this paper. We have carefully rewritten the conclusion to make it more explicit. The rewritten conclusion is as follows:
“In this paper, a new fault identification method of the diesel engine valve clearance is proposed, which is based on the improved VMD and RF. It not only can alarm the valve abnormality, but also can identify the specific fault type accurately. In the key process of nonstationary signal decomposition of this method, a novel parameters optimization strategy for VMD is carried out, which is based on reducing the energy leakage of the IMF component frequency spectrum and the minimum IMF component of power spectral entropy. Moreover, RF algorithm is employed to rank the importance of all the extracted features for the purpose of reducing the dimension and improving the computational efficiency. According to the threshold value, the features with greater importance are selected to form the feature vector. The verification results under the variable operating condition of diesel engine proved that the fault diagnosis method can effectively recognize the specific fault type of valve clearance, which average recognition accuracy reached more than 90%. In addition, this paper also makes a comparative analysis with frequently-used feature extraction and fault detection methods. The combined VMD and RF method proposed in this paper has a better identification rate for the three valve clearance faults. In the future, the valve clearance detection approach will be further studied on different engines and finally built into a real-time monitoring system.”
Point 5: Altogether, the paper needs modification to be suitable for the standards required for publication; therefore I recommend that it required to major revision.
Response 5: We are most grateful to the reviewer for many insightful and detailed comments and suggestions, which have materially improved our manuscript. Each and every comment and suggestion by the reviewer have been very carefully considered and substantial revisions have been made accordingly.
Author Response File: Author Response.pdf
Reviewer 3 Report
In this paper, a new fault identification method of the diesel engine valve clearance is proposed, which is based on the improved VMD and RF. Results are clear and well presented. In my opinion all corrections are usefull and well done. Paper can be accepted.
Author Response
Point 1:In this paper, a new fault identification method of the diesel engine valve clearance is proposed, which is based on the improved VMD and RF. Results are clear and well presented. In my opinion all corrections are usefull and well done. Paper can be accepted.
Response 1: Many thanks to the reviewer. We are very grateful that the paper can be accepted.
Round 2
Reviewer 2 Report
Dear Authors,
Please, contact the authors and inform them to resubmit the revised manuscript again, and take into consideration that the new corrections should be written using the red color font (or different color).
Best regards,
Reviewer
Author Response
Point 1: Please, contact the authors and inform them to resubmit the revised manuscript again, and take into consideration that the new corrections should be written using the red color font (or different color).
Response 1: Many thanks to the reviewer. We have reddened the modified parts of the article as required.
Round 3
Reviewer 2 Report
Dear Authors,
The Authors didn’t revise the manuscript according to my review report. Therefore, I would like to remember you with the review report that I sent last time.
Please read it carefully and revise the manuscript step by step according to the review report.
Best regards,
Reviewer
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
