AN APPROACH TO STRUCTURAL APPROXIMATION ANALYSIS BY ARTIFICIAL NEURALNETWORKS

This paper theoretically proves that a three-layer neural network can be applied to implementing exactly the function between the stresses a nd displacements and the design variables of any elastic structure bas ed on the Kolmogorov's mapping neural network existence theorem A new approach to the structural approximation analysis with the global char acteristic based on artificial neural networks is presented. The compu ter simulation experiments made by this paper show that the new approa ch is effective.