Incipient Fault Detection and Recognition of China Railway High-Speed (CRH) Suspension System Based on Probabilistic Relevant Principal Component Analysis (PRPCA) and Support Vector Machine (SVM)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This work deals with incipient fault diagnosis of a CRH suspension system. The fault diagnosis system is based on the probabilistic relevant principal component analysis and the support vector machine (SVM). The authors used simulation data from the Simpack-Matlab/Simulink co-simulation platform. After, the nonlinear PRPCA approach is used for fault detection and data preprocessing in the suspension system, and the SVM is used for fault classification. The authors used the F1-Measure index to evaluate the performance of the fault diagnosis system. The impact of the research is in the automatic control field, especially, in the fault diagnosis systems area.

 Below some comments are given for the authors.

1.      The abstract must improve, authors need to mention the main findings and limitations of the research.

2.      The introduction is section 1. Please, correct this.

3.      References used in the introduction are more than 5 years of publication. Only 37.5% of the references are new. The authors must improve the introduction section. The authors should include a conclusion of the bibliographic analysis. Also, the authors must explain the problem and why they chose this type of solution.

4.      The authors must reorganize the manuscript according to the MDPI template.

1) Introduction.

2) Material and Methods.

3) Results.

4) Discussion.

5) Conclusions

5.      Mathematical analysis seems correct. Moreover, the authors support the mathematical analysis with simulation.

6.      In section 2.3, the authors could consider including a diagram.

7.      Figure 3. Simpack-Matlab/Simulink co-simulation platform is unnecessary.

8.      The results section must be improved.

9.      The authors must give detailed information about figures 11, 12, 13, 14, 15, 16, and 17 in the results section. Preferably, the authors should present the information about the results and after, show the figure.

10.   The authors must explain the evaluation given in Table 3.

11.   Conclusions must be improved. The authors must provide the main findings of the research and the limitations.

Author Response

Thank you for the reviewer’s comments. A detailed response is provided in the attached PDF for your review.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The evaluated article deals with methods of assessing the technical condition of rolling stock suspension, in particular CRH railroad vehicles. Such assessment, if done quickly and effectively, can be essential to rail transportation safety. Thus, the subject of the research, meaning the early detection of deterioration or symptoms of suspension damage, is of great practical importance. The authors of this paper focused on simulation studies of rail vehicle models, and in this regard proposed their own modification of the PRPCA method. As the results of simulation studies show, such an upgraded method improves the accuracy and sensitivity of suspension system fault detection compared to methods described in the literature. In the paper, the authors used the resources of the Simpack-Matlab/Simulink simulation platform, where they created models and conducted tests on suspension modules. The presented work is original and it was written in an understandable way. Therefore, I recommend the work for publication - after in carrying out some of the supplements. These additions will not require re-review.

In my opinion, the weaker points of the work are:
- state of the art characterization and its relationship with the simulated model,
- the description and interpretation of the research results contained in subsection 3.2,
- too short and too general conclusion.

I suggest supplementing the introductory background with a state of the art characterization. We know from the content of the paper that the research is related to the safety and stability of the suspension system of CRH trains. So this research is about relating the simulated models to the characteristics of real rail vehicles. Hence, the reader would like to know the railcar suspension systems used in practice, the reader would also like to understand the type of defects that occur and their frequency of occurrence, evaluating the proposed method also needs information about the current suspension diagnostic methods that are used in practice.

The results of the study and their interpretation and evaluation are the most important part of the work. Therefore, I recommend - supplement subsection 3.2. with the interpretation of the results shown in each of the figures there. The results of the work should not only be included, but also described in detail, such interpretation is very important for the reader, as it allows to understand the described phenomenon more quickly and correctly. The earlier chapters contain very precise descriptions: whether the state of the knowledge, explanations of analytical methods and formulas, as well as the methodology for developing the model of the simulated rolling stock. In subsection 3.2, only drawings are inserted and a few general sentences of description are added at the end.

The Conclusions summary is too short and general. This part of the paper should give a critical evaluation of the results obtained, and highlight the strengths and weaknesses of the described solution. It would be appropriate to compare the obtained results with the results of diagnostic methods described in sources of knowledge or sources of the state of the art. It is best when such a comparison is expressed in numerical values, which makes the given results more believable.

I do not report detailed editorial comments, except for the need to complete the description of the position lterature [24] - in line 427.

Author Response

Thank you for the reviewer’s comments. A detailed response is provided in the attached PDF for your review.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The reviewer would like to thank the authors for responding to all concerns and recommendations.