Diagnosis for Vibration Fault of Steam Turbine Based on Modified Particle Swarm Optimization Support Vector Machine

Abstract:

Due to the influence of artificial factor and slow convergence of particle swarm algorithm (PSO) during parameters selection of support vector machine (SVM), this paper proposes a modified particle swarm optimization support vector machine (MPSO-SVM). A Steam turbine vibration fault diagnosis model was established and the failure data was used in fault diagnosis. The results of application show the model can get automatic optimization about the related parameters of support vector machine and achieve the ideal optimal solution globally. MPSO-SVM strategy is feasible and effective compared with traditional particle swarm optimization support vector machine (PSO-SVM) and genetic algorithm support vector machine (GA-SVM).

• Periodical: Applied Mechanics and Materials (Volumes 128-129)
• Pages: 113-116
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.128-129.113
• Online since: October 2011
• Authors: Zhi Biao Shi, Quan Gang Song, Ming Zhao Ma
• Keywords: Genetic Algorithm (GA), Particle Swarm Optimization Algorithm (PSO), Support Vector Machine (SVM), Vibration Diagnosis

References

[1] Karim S , Mojtaba K, Majid S. K, in: Fault Detection and Diagnosis of An Industrial Steam Turbine Using Fusion of SVM and ANFIS Classifiers[J]. Energy, 2010, (35): 5472-5482.

DOI: 10.1016/j.energy.2010.06.001

[2] Eslam P, Siavash R, in: Application of Genetic Algorithm-Support Vector Machine (GA-SVM) for prediction of BK-channels activity[J]. European Journal of Medicinal Chemistry, 2009, (44): 5023-5028.

DOI: 10.1016/j.ejmech.2009.09.006

[3] Vapnik V. N, in : The Nature of Statistical Learning Theory [M]. New York: Springer-Verlag, (1995).

[4] Zhao Gang, Jiang DongXiang, etc, in: Data Mining for Fault Diagnosis and Machine Learning for Rotating Machinery[J]. Key Engineering Materials, 2005, (293-294): 175-182.

DOI: 10.4028/www.scientific.net/kem.293-294.175

[5] Vahid R, Ata E, Reza G, in: Application of The PSO-SVM Model for Recognition of Control Chart Patterns[J]. ISA Transactions, 2010, (49): 577-586.

DOI: 10.1016/j.isatra.2010.06.005

[6] Ge Zhi Hua, in : Study on Vibration Fault of Turbine Shafts Based on Non-linear Theory (Ph.D., North China Electric Power University, China 2000): 36-37.

[7] Chu Yan, in : Support Vector Machines and Its Application in The Thermal Power Engineering Field (MS., North China Electric Power University, China 2006): 31-32.