Optimal influence cover for an element free Galerkin MFree method based on artificial neural network

Abstract

The present investigation presents an efficient meshless method based on the weak form of an element-free-Galerkin method. The formulation of the numerical solution was conducted using an artificial neural network (ANN) approach to compute the optimal number of nodes in the influence domain for each point of interest. The numerical results using the ANN model were tested and compared with different approaches in the literature. Results show a reduction in the computational cost and an enhancement in an error criterion of up to 11%.