DocumentCode :
1521732
Title :
Algorithms for Fast Simulation of Transformer Windings for Diagnostic Testing of Power Transformers
Author :
Singh, Arvind ; Marti, Jose R. ; Srivastava, Krishan D.
Author_Institution :
Centre for Energy Syst., Univ. of Trinidad & Tobago, Couva, Trinidad and Tobago
Volume :
25
Issue :
3
fYear :
2010
fDate :
7/1/2010 12:00:00 AM
Firstpage :
1564
Lastpage :
1572
Abstract :
Accurate high-frequency simulation of power transformer windings often involves detailed turn-by-turn modeling of the winding structure. The most detailed models currently in existence use multiphase transmission systems with frequency-dependent losses. The conventional Electromagnetic Transients Program method of solving these systems involves eigenvalue decompositions for hyperbolic corrections as well as matrix inversions, both of which are prohibitively computationally expensive. This paper presents three optimized algorithms for improving the simulation speed of the most common circuit models by taking advantage of useful mathematical identities. A diakoptic method of splitting the system into series and shunt elements to exploit the sparsity of the shunt admittance system is also presented. The result is a significant speed increase, making the simulation of large systems more feasible and minimizing the need for circuit-reduction techniques.
Keywords :
EMTP; eigenvalues and eigenfunctions; power transformer testing; transformer windings; circuit-reduction techniques; diakoptic method; eigenvalue decompositions; electromagnetic transient program method; fast simulation algorithm; frequency-dependent losses; high-frequency simulation; hyperbolic corrections; matrix inversions; multiphase transmission systems; power transformer diagnostic testing; power transformer winding structure; shunt admittance system; Circuit simulation; EMTP; Eigenvalues and eigenfunctions; Frequency; Matrix decomposition; Power system modeling; Power transformers; Propagation losses; Testing; Windings; Diakoptics; fast simulation; transformer modeling;
fLanguage :
English
Journal_Title :
Power Delivery, IEEE Transactions on
Publisher :
ieee
ISSN :
0885-8977
Type :
jour
DOI :
10.1109/TPWRD.2009.2037630
Filename :
5491379
Link To Document :
بازگشت