Development of ANN-based preliminary structural design systems for cable-stayed bridges

Abstract

The development of computer-supporting systems in early design stage has been suffering from the difficulty in computerizing the information and knowledge because it lacks formality of representation of designers' experience and intuition. To overcome this problem, the concept of artificial neural network (ANN) was adopted to develop preliminary design system for cable-stayed bridges in this paper. Considering that relatively little is known in early design stages and design information evolves as design proceeds, multi-level networks for progressive structural design are proposed. This approach has been proved to have the ability of learning design experience and providing practical design alternatives. The prototype of preliminary structural design system has been implemented as a part of integrated design system to show the process of the ANN-based preliminary structural design proposed in this paper. The proposed methodology is hoped to help facilitate more effective preliminary design toward an integrated computer-supporting design system.