Faults and Diagnosis Methods of Permanent Magnet Synchronous Motors: A Review

Round  1

Reviewer 1 Report

The paper proposes a review of faults and diagnosis methods of Permanent Magnet Synchronous Motors (PMSM). After an introduction reminding the large application field of PMSMs, the paper describes the faults that may occur in this motor. The paper is based on 108 references to research works.

However several affirmations are questionable:

 - The incorrect connection of motor windings is presented as a fault of the motor. It is a strange denomination for an error of the operator in charge of the motor connection. In my opinion a fault is an event that happens on a working motor, not an error of the operator installing the machine.

 - The inter-turn short circuit is an actual a fault. The consequences are described but not the causes. A short paragraph should be added for detailing the causes that may cerate such turn-to turn faults.

 - In the description of mechanical faults, what do you mean by “missing of the bolts”? In my opinion it is not a motor fault but an error of the operator.

 - In section 3.1, you describe only secondary phenomena, which are the increase of several harmonics. In my opinion the main consequence of the turn-to turn is a local hot spot that can burn the wire insulation. There is not a word on this major phenomenon; a short paragraph should be added on the hot spot temperatures and the time constants of the associated transients.

 - Section 4.1 gives with very basic equations, which are valid only for linear operations and for a smooth stator, neglecting many effects. This part should be removed because the equations are not really used in the paper.

 - In section 4.2, you describe several signal analysis methods. There is neither analysis nor comparison of these mathematical approaches.

 - Section 5 dealing with discussion is too short.

 The paper gives only a superficial description of the fault diagnosis methods. The reader cannot get any indication useful for applications. The section “discussion” should be rewritten in details with a comparative analysis of the methods described above.

Author Response

 Author Response File: Author Response.pdf

Reviewer 2 Report

The review is interesting, but needs improvements:

- English needs improvements;

- the structure of the paper should be revised;

- Section 3 is discussing two different signal based techniques, thus the section title “fault performance” is not corresponding to the intended content. Authors should check this detail.

- Section 3 is discussing different signal based techniques which should be included/or presented after the section “4. Fault diagnosis methods”

- Steps of the Fault diagnosis flowchart of PMSM presented in Figure 1 should be numbered and commented step by step for more clarity for the readers.

- According to Figure 1, all diagnosis techniques are derived from signal based techniques. Can the authors clarify the proposed vision?

-  Equations 8-9 are well known equations and not necessary for the discussion. Can the authors substitute them by a more focused development or bloc schemes oriented to the issue of model based techniques?

- The subsection “Signal processing methods” is related to signal based techniques. Section 3 and subsection 4.2 could be fused for a new section or subsection under signal based techniques.

- An important issue related to techniques performances against the operating conditions, and mainly transient operating conditions, which is a hot topic for modern fault diagnosis techniques is completely absent? Authors should at least discuss this issue for the different diagnosis approaches.

 - The above issue should be discussed mainly for STFT, WT, HT, … showing their advantages and limits in transient operating conditions;

- Variables of equations (11)-(14) should be clearly defined, then, performance discussion of each technic should be discussed starting from the concerned parameters

Author Response

Reviewer 2

Comments:

The review is interesting, but needs improvements:

- English needs improvements;

Response: Many thanks for your careful check! We are sorry for our negligence. In the revised manuscript, we corrected the English grammar mistakes and revised the wording.

- the structure of the paper should be revised;

- Section 3 is discussing two different signal based techniques, thus the section title “fault performance” is not corresponding to the intended content. Authors should check this detail.

- Section 3 is discussing different signal based techniques which should be included/or presented after the section “4. Fault diagnosis methods”

Response: Many thanks for your comments! In the revised manuscript, we have moved the signal features in Section 3 into Section 4, which is Section 3 now, after the introduction of the frequency domain method.

- Steps of the Fault diagnosis flowchart of PMSM presented in Figure 1 should be numbered and commented step by step for more clarity for the readers.

Response: Many thanks for your comments! In the revised manuscript, we have added more descriptions of each step in the flowchart.

- According to Figure 1, all diagnosis techniques are derived from signal based techniques. Can the authors clarify the proposed vision?

Response: Many thanks for your comments! All the fault diagnosis methods, in order to determine the actual motor fault type, need its signal data. The difference is that some methods are to compare these data with the data of math model, some methods are to process these signal data to find the features, and some methods are to apply the data to the training of machine learning model. The motor signal data is the base of fault diagnosis. In the revised manuscript, we added some statement about it.

- Equations 8-9 are well known equations and not necessary for the discussion. Can the authors substitute them by a more focused development or bloc schemes oriented to the issue of model based techniques?

Response: Many thanks for your comments! In the revised manuscript, the basic equations were removed. And we have added more introductions and evaluations on the development of MEC and FEM models.

- The subsection “Signal processing methods” is related to signal based techniques. Section 3 and subsection 4.2 could be fused for a new section or subsection under signal based techniques.

Response: Many thanks for your comments! In the revised manuscript, we have moved the signal features in Section 3 into Section 4, which is Section 3 now, after the introduction of the frequency domain method.

- An important issue related to techniques performances against the operating conditions, and mainly transient operating conditions, which is a hot topic for modern fault diagnosis techniques is completely absent? Authors should at least discuss this issue for the different diagnosis approaches.

 - The above issue should be discussed mainly for STFT, WT, HT, … showing their advantages and limits in transient operating conditions;

Response: Many thanks for your comments! We have added more description and comparison in Section 3.2, including their applicability to transient signals.

- Variables of equations (11)-(14) should be clearly defined, then, performance discussion of each technic should be discussed starting from the concerned parameters

Response: Many thanks for your careful check! We added more description of the formulas and more detailed evaluations of the time-frequency analysis methods based on them.

Author Response File: Author Response.pdf

Round  2

Reviewer 1 Report

The paper has been significantly improved. It is now a very interinting review on fault detection on PMSM, with many sound references.

Author Response

thanks for your postive comments,thanks!

Reviewer 2 Report

The paper has been improved and authors have clearly adressed all my previous concerns.

Author Response

thanks for your postive comments,  English language and style are  spell check