Compressor Surge Detection Based on Support Vector Data Description

Abstract:

Axial flow compressors work as an indispensable device in industry fields. Surge is a phenomenon of aerodynamic instability, which characterized by disruption of flow. When a compressor works in surge state, the vibration is so intense that it may causes accidents. Detecting surge timely and accurately not only insure safety of compressors but also is a key of active control of aerodynamic instability. Support vector data description (SVDD) is a one-class classification method developed based on the theory of support vector machine (SVM). In this paper, we introduce SVDD into the field of compressor surge detection. It demonstrates that SVDD method can give a warning far ahead of surge.

• Periodical: Applied Mechanics and Materials (Volumes 152-154)
• Pages: 1545-1549
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.152-154.1545
• Online since: January 2012
• Authors: Chang Zheng Li, Yong Lei
• Keywords: Aerodynamic Instability, Axial Flow Compressor, Signal Identification, Support Vector Data Description, Surge

References

[1] H. G. Zhang, W. L. Chu, Y. H. Wu, and et al, Numerical Investigation of the Flow Mechanism of Compressor Stall Delay through End Wall Self Recirculation, J. Propulsion and Technology, 2009, Vol. 30, No. 2, pp.202-207.

[2] L. Li, and C. Q. Nie, On-line Measurement and Stall Control by DSP in Single-stage Axial Compressor, Acta Aeronautica et Astronautica Sinica, 2009, Vol. 30, No. 5, pp.787-793.

[3] M Dhingra, J Armor, Y Neumeier, and et al, Compressor Surge: A Limit Detection and Avoidance Problem, AIAA Guidance, Naviagtion and Control Conf. and Exhibit, Aug. 2005, San Francisco, California.

DOI: 10.2514/6.2005-6449

[4] M Dhingra, Y Liu, Y Neumeier, and et al, Comressor Stability Detection on a Turboshaft Engine with a Hybird Axi-centrifugal Compressor, 43rd AIAA/ASME/SAE/ASEE Joint Propulsion Conf. and Exhibit, July 2007, Cincinnati, OH.

DOI: 10.2514/6.2007-5058

[5] C A Brown, S Sawyer, W Oakes, and et al, Wavelet Based Analysis of Rotating Stall and Surge in a High Speed Centrifugal Compressor, 38th AIAA/ASME/SAE/ASEE Joint Propulsion Conf. and Exhibit, July 2002, Indianapolis IN.

DOI: 10.2514/6.2002-4080

[6] J. Xu, D. Y. Shi, Y. J. Zhang, and et al, Model of IDS Based on SVDD and Cluster Algorithm, Control and Decision, 2010, Vol. 25, No. 3, pp.441-444.

[7] M. Q. Pan, S. X. Qian, L. Y. Lei, and et al, Support Vector Data Description with Model Selection for Condition Monitoring, Proceeding of the 4th International Conference on Machine Learning and Cybernetics, 18-21 August, 2005, Guangzhou, China.

DOI: 10.1109/icmlc.2005.1527696

[8] Y. H. Liu, Y. C. Liu, and Y. J. Chen, Fast Support Vector Data Descriptions for Novelty Detection, IEEE Transactions on Neural Networks, 2010, Vol. 21, No. 8, pp.1296-1313.

DOI: 10.1109/tnn.2010.2053853

[9] J. Luo, B. Li, C. Q. Wu, and Y. H. Pan, A Fast SVDD Algorithm Based on Decomposition and Combination for Fault Detection, 8th IEEE International Conference on Control and Automation, 9-11 June, 2010, Xiamen, China.

DOI: 10.1109/icca.2010.5524160

[10] Rolls-Royce plc, The Jet Engine, The Technical Publications Department, Rolls-Royce plc, Derby, England, (1996).

DOI: 10.3940/rina.ws.2005.16

[11] H. Zhang, and Q. H. Pu, Aerodynamic Unsteadiness Detection in Compressor Based on Wavelet Analysis, Gas Turbine Experiment and Research, 2005, Vol. 18, pp.38-40.