Abstract
The application of neural networks to cold-formed steel design is considered. Cold-formed steel design is more complex than hot-rolled steel design because of the large number of different section profiles, and traditional computing tools are not well-suited for dealing with this. The flexibility of neural networks makes them a suitable alternative. The paper describes the use of neural networks to predict the failure load of cold-formed sections, and the results are in good agreement with results from design codes.
Original language | English |
---|---|
Title of host publication | Computational engineering using metaphors from nature |
Editors | B.H.V. Topping |
Place of Publication | Edinburgh |
Publisher | Civil-Comp Press |
Pages | 37-43 |
Number of pages | 7 |
Volume | 64 |
ISBN (Print) | 0948749660 |
DOIs | |
Publication status | Published - 2000 |
Event | 5th International Conference on Computational Structures Technology/2nd International Conference on Engineering Computational Technology - LEUVEN, Belgium Duration: 6 Sept 2000 → 8 Sept 2000 http://www.civil-comp.com/conf/prog2000.htm |
Conference
Conference | 5th International Conference on Computational Structures Technology/2nd International Conference on Engineering Computational Technology |
---|---|
Country/Territory | Belgium |
City | LEUVEN |
Period | 6/09/00 → 8/09/00 |
Internet address |