Neural networks in mechanics of structures and materials - new results andprospects of applications

Basic ideas of back-propagation neural networks (BPNNs) are presented in sh ort. Then BPNN applications in analysis of the following problems are discu ssed: (1) bending analysis of elastoplastic beams, (2) elastoplastic plane stress problem, (3) estimation of fundamental vibration periods of real bui ldings, (4) detection of damage in a steel beam. (5) identification of load s applied to an elastoplastic beam. Regularization neural network is briefl y discussed and its application to estimation of concrete fatigue durabilit y it shown. A modified Hopfield network is used to the analysis of an elast ic angular plate with unilateral constraints. In the end some conclusions a nd prospects of neurocomputing applications are pointed out. (C) 2001 Elsev ier Science Ltd. All rights reserved.