Finite element analysis based Hopfield neural network model for solving nonlinear electromagnetic field problems

Abstract

Based on the finite element method for solving nonlinear magnetic field problems, a locally connected Hopfield network is constructed by taking the energy functional of the boundary value problem as the computing energy function of this neural network. The fact that the steady point of the dynamic system must correspond to the numerical solution of the finite element method is explained explicitly. The numerical results show the effectiveness and correctness of the proposed method.