KEPCO Journal on Electric Power and Energy

Korea Electric Power Corporation (한국전력공사)
권호 표지
DOI QR코드

DOI QR Code

Fault Diagnosis Method based on Feature Residual Values for Industrial Rotor Machines

  • Kim, Donghwan (KEPCO Research Institute, Korea Electric Power Corporation) ;
  • Kim, Younhwan (KEPCO Research Institute, Korea Electric Power Corporation) ;
  • Jung, Joon-Ha (Seoul National University, Department of Mechanical and Aeropsace Engineering) ;
  • Sohn, Seokman (KEPCO Research Institute, Korea Electric Power Corporation)
  • Received : 2018.07.03
  • Accepted : 2018.11.12
  • Published : 2018.12.30
Download PDF
⟨ Previous Next ⟩

Abstract

Downtime and malfunction of industrial rotor machines represents a crucial cost burden and productivity loss. Fault diagnosis of this equipment has recently been carried out to detect their fault(s) and cause(s) by using fault classification methods. However, these methods are of limited use in detecting rotor faults because of their hypersensitivity to unexpected and different equipment conditions individually. These limitations tend to affect the accuracy of fault classification since fault-related features calculated from vibration signal are moved to other regions or changed. To improve the limited diagnosis accuracy of existing methods, we propose a new approach for fault diagnosis of rotor machines based on the model generated by supervised learning. Our work is based on feature residual values from vibration signals as fault indices. Our diagnostic model is a robust and flexible process that, once learned from historical data only one time, allows it to apply to different target systems without optimization of algorithms. The performance of the proposed method was evaluated by comparing its results with conventional methods for fault diagnosis of rotor machines. The experimental results show that the proposed method can be used to achieve better fault diagnosis, even when applied to systems with different normal-state signals, scales, and structures, without tuning or the use of a complementary algorithm. The effectiveness of the method was assessed by simulation using various rotor machine models.

Keywords

JROGG5_2018_v4n2_89_f0001.png 이미지

Fig. 1. Fault diagnosis framework of rotor machinery based on feature residual values classification approach.

JROGG5_2018_v4n2_89_f0001.png 이미지

Fig. 1. Fault diagnosis framework of rotor machinery based on feature residual values classification approach.

JROGG5_2018_v4n2_89_f0002.png 이미지

Fig. 2. Schematic of the procedure for generating the feature residual values.

JROGG5_2018_v4n2_89_f0002.png 이미지

Fig. 2. Schematic of the procedure for generating the feature residual values.

JROGG5_2018_v4n2_89_f0003.png 이미지

Fig. 3. First experimental setup: (a‐1) RK‐4 rotor kit with (a‐2) displacement sensor and (a‐3) acceleration sensor, the fault imbedding device for (a‐4) mass unbalance, (a‐5) disk rubbing, (a‐6) misalignment and (a‐7) oil whirl. Second experimental setup: (b‐1) small‐scale steam turbine prototype with (b‐2) measurement equipment, (b‐3) bearing and two shafts, the fault imbedding device for (b‐4) mass unbalance, (b‐5) disk rubbing, (b‐6) shaft rubbing, and (b‐7) misalignment. Third experimental setup: (c‐1) industrial steam turbine with (c‐2) shaft and blades and (c‐3) a bearing.

JROGG5_2018_v4n2_89_f0003.png 이미지

Fig. 3. First experimental setup: (a‐1) RK‐4 rotor kit with (a‐2) displacement sensor and (a‐3) acceleration sensor, the fault imbedding device for (a‐4) mass unbalance, (a‐5) disk rubbing, (a‐6) misalignment and (a‐7) oil whirl. Second experimental setup: (b‐1) small‐scale steam turbine prototype with (b‐2) measurement equipment, (b‐3) bearing and two shafts, the fault imbedding device for (b‐4) mass unbalance, (b‐5) disk rubbing, (b‐6) shaft rubbing, and (b‐7) misalignment. Third experimental setup: (c‐1) industrial steam turbine with (c‐2) shaft and blades and (c‐3) a bearing.

JROGG5_2018_v4n2_89_f0004.png 이미지

Fig. 4. Original vibration signal (time‐domain) in four conditions and their corresponding various feature residual trends. (a) normal‐state, (b) misalignment, (c) rub, (d) crack, (e) 1s average of skewness (good case), (f) 1s deviation of crest factor (good case), (g) 1s average of max value (good case), (h) 1s deviation of shape factor (bad case).

JROGG5_2018_v4n2_89_f0004.png 이미지

Fig. 4. Original vibration signal (time‐domain) in four conditions and their corresponding various feature residual trends. (a) normal‐state, (b) misalignment, (c) rub, (d) crack, (e) 1s average of skewness (good case), (f) 1s deviation of crest factor (good case), (g) 1s average of max value (good case), (h) 1s deviation of shape factor (bad case).

JROGG5_2018_v4n2_89_f0005.png 이미지

Fig. 5. Comparisons of fault identification accuracy rates of RPC based classifier and common FDA classifier for different experimental cases: the classification results obtained using (a) different baselines for the normal state and (b) identical training data and different target systems.

JROGG5_2018_v4n2_89_f0005.png 이미지

Fig. 5. Comparisons of fault identification accuracy rates of RPC based classifier and common FDA classifier for different experimental cases: the classification results obtained using (a) different baselines for the normal state and (b) identical training data and different target systems.

JROGG5_2018_v4n2_89_f0006.png 이미지

Fig. 6. The example process of generating feature residual values on real data set from vibration sensors. (include normal and faulty vibration data).

JROGG5_2018_v4n2_89_f0006.png 이미지

Fig. 6. The example process of generating feature residual values on real data set from vibration sensors. (include normal and faulty vibration data).

JROGG5_2018_v4n2_89_f0007.png 이미지

Fig. 7. Visualization of the classification obtained by the RPC based diagnosis algorithm by changing the normal state of the system using four selected features.

JROGG5_2018_v4n2_89_f0007.png 이미지

Fig. 7. Visualization of the classification obtained by the RPC based diagnosis algorithm by changing the normal state of the system using four selected features.

Table 1. Explanations of training and testing data sets for each state of the rotor machine.

JROGG5_2018_v4n2_89_t0001.png 이미지

Table 1. Explanations of training and testing data sets for each state of the rotor machine.

JROGG5_2018_v4n2_89_t0001.png 이미지

Table 2. Calculation sample data obtained from vibration sensors: each table includes fault data (left), estimated values of measured data (center) and the feature residual values obtained by RPC based algorithm (right).

JROGG5_2018_v4n2_89_t0002.png 이미지

Table 2. Calculation sample data obtained from vibration sensors: each table includes fault data (left), estimated values of measured data (center) and the feature residual values obtained by RPC based algorithm (right).

JROGG5_2018_v4n2_89_t0002.png 이미지

Table 3. Fault diagnosis results under various classifiers with feature residuals and raw feature values. (The test are conducted by representative experiments from No.2, 4, 5 and 7)

JROGG5_2018_v4n2_89_t0003.png 이미지

Table 3. Fault diagnosis results under various classifiers with feature residuals and raw feature values. (The test are conducted by representative experiments from No.2, 4, 5 and 7)

JROGG5_2018_v4n2_89_t0003.png 이미지

References

  1. Edwards, S., Lees, A., Friswell, M., "Fault diagnosis of rotating machinery." Shock and Vibration Digest, 30, (1), 1998, 4-13. https://doi.org/10.1177/058310249803000102
  2. Jalan, A. K., Mohanty, A. R., "Model based fault diagnosis of a rotor-bearing system for misalignment and unbalance under steady‐state condition." J. Sound Vibr., 327, (3-5), 2009, 604-622. https://doi.org/10.1016/j.jsv.2009.07.014
  3. Sekhar, A. S., "Crack identification in a rotor system: a model-based approach." J. Sound Vibr., 270, (4-5), 2004, 887-902. https://doi.org/10.1016/S0022-460X(03)00637-0
  4. Mahadevan, S., Shah, S., "In Plant Wide Fault Identification using One‐class Support Vector Machines." Fault Detection, Supervision and Safety of Technical Processes, 2009, pp 1013-1018.
  5. Li, C. J., Shin, H., "In Tracking Bearing Spall Severity Through Inverse Modeling." ASME 2004 International Mechanical Engineering Congress and Exposition, Anaheim, California, USA, 2004, pp 49‐54.
  6. Juricic, D., Moseler, O., Rakar, A., "Model‐based condition monitoring of an actuator system driven by a brushless DC motor. Control Engineering Practice." (5), 2009, 545-554.
  7. Angelov, P., Giglio, V., Guardiola, C., Lughofer, E., Lujan, J. M., "An approach to model‐based fault detection in industrial measurement systems with application to engine test benches." Measurement Science and Technology, 17, (7), 2006, 1809. https://doi.org/10.1088/0957-0233/17/7/020
  8. de Araujo Ribeiro, R. L., Jacobina, C. B., Cabral da Silva, E. R., Lima, A. M. N., "Fault detection of open‐switch damage in voltage‐fed PWM motor drive systems." Power Electronics, IEEE Transactions, 18, (2), 2003, 587-593. https://doi.org/10.1109/TPEL.2003.809351
  9. Yang, Q., "Model‐based and data driven fault diagnosis methods with applications to process monitoring." Case Western Reserve University, 2004.
  10. Piramanrique, M., Francisco, R., Sofrony Esmeral, J., "Data driven fault detection and isolation: a wind turbine scenario." Tecnura, 19, (44), 2015, 71-82. https://doi.org/10.14483/udistrital.jour.tecnura.2015.2.a05
  11. Yuan, S.‐F., Chu, F.‐L., "Support vector machines‐based fault diagnosis for turbo‐pump rotor." Mechanical Systems and Signal Processing, 20, (4), 2006, 939-952. https://doi.org/10.1016/j.ymssp.2005.09.006
  12. Santos, P., Villa, L., Renones, A., Bustillo, A., Maudes, J., "An SVM‐Based Solution for Fault Detection in Wind Turbines." Sensors, 15, (3), 2015, 5627. https://doi.org/10.3390/s150305627
  13. Allgood, G. O., Upadhyaya, B. R., "In Model‐based high-frequency matched filter arcing diagnostic system based on principal component analysis (PCA) clustering." AeroSense International Society for Optics and Photonics, 2000, pp 430-440.
  14. Li, P., Li, X. J., Jiang, L. L., Yang, D. L., "Fault Diagnosis for Motor Rotor Based on KPCA‐SVM." applied mechanics and materials, 2011, 143-144, 680-684. https://doi.org/10.4028/www.scientific.net/AMM.143-144.680
  15. Li, Y., Wang, Z., Yuan, J., "On‐line Fault Detection Using SVM‐based Dynamic MPLS for Batch Processes." Chinese Journal of Chemical Engineering, 14, (6), 2006, 754-758. https://doi.org/10.1016/S1004-9541(07)60007-X
  16. Zhu, H. Q., Fang, R. M., Peng, C. Q., "The Application of ICA‐SVM Method for Identifying Multiple Faults in Asynchronous Motors." Applied Mechanics and Materials, 483, 2013, 405-408. https://doi.org/10.4028/www.scientific.net/AMM.483.405
  17. Zhao, Q., "The Study on Rotating Machinery Early Fault Diagnosis based on Principal Component Analysis and Fuzzy C-means Algorithm." Journal of Software, 8, (3), 2013, 709-715.
  18. Tao, X., Lu, C., Lu, C., Wang, Z., "An approach to performance assessment and fault diagnosis for rotating machinery equipment." EURASIP J. Adv. Signal Process, (1), 2013, 1-16.
  19. Ma, J., Jiang, J., "Applications of fault detection and diagnosis methods in nuclear power plants: A review." Progress in Nuclear Energy, 53, (3), 2011, 255-266. https://doi.org/10.1016/j.pnucene.2010.12.001
  20. Yin, S., Wang, G., Karimi, H. R., "Data‐driven design of robust fault detection system for wind turbines." Mechatronics 24, (4), 2014, 298-306. https://doi.org/10.1016/j.mechatronics.2013.11.009
  21. Technical Review of On‐line Monitoring Techniques for Performance Assessment. Vol. 3.
  22. Liu, Z., Zuo, M. J., Xu, H., "Feature ranking for support vector machine classification and its application to machinery fault diagnosis." Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science 2012, 2077-2089.