Title :
Mesh generation of FEM by ANN on iron — Core transformer
Author :
Srasuay, K. ; Chumthong, A. ; Ruangsinchaiwanich, S.
Author_Institution :
Dept. of Electr. & Comput. Eng., Naresuan Univ., Muang, Thailand
Abstract :
Artificial neural network (ANN) has emerged as a potential tool for processing many applications because it is capable of recognizing patterns. This study is proposed to apply feed-forward neural network to generate mesh models in finite element method for analysis of iron-core transformer. The results of the proposed neural network for mesh generation are compared with those of Delaunay and random method. Mesh generation from the proposed method shows satisfactory results. Results from error norm approximation and finite element calculation are consistent.
Keywords :
approximation theory; feedforward neural nets; mesh generation; pattern recognition; random processes; transformer cores; ANN; Delaunay method; FEM; artificial neural network; error norm approximation; feedforward neural network; finite element method; iron-core transformer; mesh generation; pattern recognition; random method; Analytical models; Approximation methods; Artificial neural networks; Finite element methods; Magnetic flux; Mesh generation; Solid modeling; ANN; Error Norm Approximation; FEM;
Conference_Titel :
Electrical Machines and Systems (ICEMS), 2010 International Conference on
Conference_Location :
Incheon
Print_ISBN :
978-1-4244-7720-3