FINITE-ELEMENT SOLUTIONS WITH FEEDBACK NETWORK MECHANISM THROUGH DIRECT MINIMIZATION OF ENERGY FUNCTIONALS

In this paper, a new algorithm for solving a minimization problem of t he functional of a variational principle was proposed by focusing on t he state transition process of the interconnected neural network. A ne w concept of neuron called 'feedback neuron' was introduced into the c onventional interconnected neural network. This method was readily app lied to the finite element analyses of incompressible viscous flows. F urthermore, favourable feature of the neural network for parallel plat form was explored on a massively parallel computer AP1000.