Fault Diagnosis Based on Particle Swarm Fuzzy Clustering Algorithm

Abstract:

Fuzzy c-means clustering algorithm (FCM) is sensitive to noise and less effective when handling high dimensional data set. Given that particle swarm optimization algorithm (PSO) has strong global search capability and efficient performance, a new PSO based fuzzy clustering algorithm is proposed. Particles in the new algorithm are encoded by membership in FCM. The new algorithm adopts a new strategy to meet the constraints of FCM, so as to optimize the clustering effect of FCM. Finally, this algorithm is applied to motor fault diagnosis. Experiment shows that the new algorithm made up for the shortcomings of FCM, improved the efficiency and accuracy of clustering and bettered fault diagnosis results.

• Periodical: Applied Mechanics and Materials (Volumes 63-64)
• Pages: 111-114
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.63-64.111
• Online since: June 2011
• Authors: Zhi Xiao Wang, Qiang Niu
• Keywords: Clustering, Fault Diagnosis, Fuzzy C-Means (FCM), Particle Swarm Algorithm

References

[1] Zhu Wen , Hou Beipin. Research on application of wavelet packet and BP neural network in fault diagnose of electromotor[J]．Control Theorems & Their Applications , 2006，28 (1): 10 – 11.

[2] Fu Guangjie, Li Yunpeng, Yang Xiuju. Research for Motor Failure Diagnosis Based on Improvement Particle Swarm Optimization Algorithm. Science Technology and Engineering, 2010, 10(4): 1001-1005.

[3] Wu Huiying, Shuai Renjun．Fuzzy set analysis of turbinerotator vibration malfunction[J ]．Journal of Shanghai Jiaotong University, 1999, 33 (8): 1051 – 1054.

[4] Zhang Bide, Pan Ling. New Approach to Fuzzy Clustering Analysis in Multiple-faults Diagnosis of a Turbogenerator Set. Journal of Xihua University(Natural Science Edition), 2006, 25 (3): 8- 10.

[5] Chen Shu. Researches on Mathematical Models for Data Clustering and Applications [D]. Wuxi, Jiangnan University, (2007).

[6] Yang Wei, Li Qiqiang. Survey on Particle Swarm Optimization Algorithm. Engineering Science, 2004, 6 (5): 87-94.

[7] WANG JW, LI HN. Summary of particle swarm optimization algorithm. Modern Computer, 2009, 301: 22-27.

[8] RUNKLER TA, KATZ C. Fuzzy clustering by particle swarm optimization[C] Proc of the IEEE International Conference on Fuzzy Sys2tems. 2006：6012608.

DOI: 10.1109/fuzzy.2006.1681773

[9] Shi Dingding, Pan Hongxia. Ant Colony Algorithm Application to the Fault Diagnosis of Motor. Large Electric Machine and Hydraulic Turbine, 2009, (1): 26-30.