Parameter Identification for Power Transformer Using Zooming Genetic Algorithms

Abstract:

Transfer functions of power transformer can be used for detection and localization of partial discharges. There is no access to sectional windings when transformer insulation has been impregnated into oil, so the parameter identification of transformer windings is difficult without the knowledge of sectional winding voltages. In this paper, genetic algorithms (GAs) based approach are proposed to identify the parameters of transformer windings. An impulse signal was injected into the bushings of transformer, then the responses at neutral terminals were measured, and thee acquired data were used to identify the parameters using GAs. To accelerate GAs’ convergence, a zooming operator was introduced in genetic programming process. The experimental results show that the proposed ZGA approach is effective.

• Periodical: Advanced Materials Research (Volumes 860-863)
• Pages: 2149-2152
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.860-863.2149
• Online since: December 2013
• Authors: Hou Quan Zhu, Rui Ming Fang, Chang Qing Peng
• Keywords: Genetic Algorithm (GA), Parameter Identification, Partial Discharge (PD), Power Transformer

References

[1] H. W. Barry: IEEE Electrical Insulation Magazine, vol. 17- 3(2001), p.16.

[2] P. Werle, A. Akbari, H. Boris, E. Gockenbbach, in: Localization and evaluation of partial discharges power transformers using sectional winding transfer functions, Proc. 12th Int'l Symp. High Volt. Eng. (2001).

[3] A. Akbari, P. Werle, H. Borsi, and E. Gockenbach, in: High frequency transformer model for computation of sectional winding transfer functions used for partial discharge localization, Proc. 12th Int'l Symp. High Volt. Eng. (2001).

[4] A. Akbari, P. Werle, H. Boris, E. Gockenbbach: IEEE Electrical Insulation Magazine, vol. 18-5(2002), p.22.

[5] D.E. Goldberg: Genetic algorithms in search, optimization, and machine learning( Addison-Wesley, USA, 1989).