Title :
A finite element mesh generator based on an adaptive neural network
Author_Institution :
Dipt. Elettrico, Elettronico e Sistemistico, Catania Univ., Italy
fDate :
9/1/1998 12:00:00 AM
Abstract :
A finite element mesh generator based on an artificial neural network is described. The generator grows an initial mesh of triangles up to a user-selected number of nodes. The mesh growth is driven by a node probability density function, which is obtained from an error estimation for the solution of the problem with the initial mesh. Good-accuracy solutions are obtained with a lower computational effort
Keywords :
adaptive systems; electric fields; mesh generation; neural nets; probability; adaptive neural network; computational effort; electric fields; error estimation; finite element mesh generator; mesh growth; node probability density function; Adaptive mesh refinement; Adaptive systems; Artificial neural networks; Error analysis; Finite element methods; Iterative methods; Mesh generation; Neural networks; Probability density function;
Journal_Title :
Magnetics, IEEE Transactions on
