A neural network for structural re-analysis

The need for a computationally efficient method of structural re-analysis i s long-standing. In general traditional methods of analysis have not proved suitable for the purpose. Recent progress in neural computing technology h as provided a more suitable background from which to develop a general iter ative re-analysis method. A simple neural network architecture conveniently accommodates design chang es in geometry, element/material properties, topology, applied loading, sup ports and is therefore suitable for insertion in a combined design/reanalys is computing environment. The weight in the neural network represent structural flexibilities and aft er training a set of weights corresponding to the initial state of the stru cture, design changes are made as required the network then continues train ing from the previous state. Illustrative examples demonstrate that the network approach, whilst conside rably slower than band-matrix processing, offers considerable advantage in convenience for the designer. (C) 1999 Elsevier Science Ltd. All rights res erved.