A method employing ideas from applied artificial intelligence is proposed for use in design optimization. Knowledge of global information is incorporated in several rule types concerning constraint activity, redundancy, and dominance. The rules are organized in a production system that makes deductions about possible (globally) active sets. Two design examples demonstrate the strategy. Similar ideas may be used in other areas of the optimization process, particularly in design applications.

This content is only available via PDF.
You do not currently have access to this content.