Recent studies at the University of Waterloo have developed adaptive computing techniques to establish Pareto-optimal performance trade-off surfaces for the conceptual structural and architectural design of office buildings. Taking into account the structural, mechanical and electrical requirements, the land, erection and maintenance costs, and the quality of occupied space for a specific building project, a multi-criteria genetic algorithm is applied to find the Pareto-optimal surface corresponding to the three objective criteria of minimum capital cost, minimum life-cycle cost and maximum revenue income. Each point on the surface represents a feasible conceptual design of the building having the property that no other design point on the surface dominates it for all three objective criteria.

The Pareto-optimal surfaces for building conceptual design provide useful information concerning the performance trade-off between the different objective criteria. However, the number of points on any one surface is generally so large that the selection of one or a few conceptual designs as being more desirable than the other designs is still a daunting task for the designer. The purpose of the present paper is to investigate means of further filtering the Pareto-optimal results to arrive at a limited number of conceptual designs that are best according to one or more composite objective criteria. For instance, designs that are likely to be the most profitable over a fixed time period even though their initial capital costs, and perhaps their life-cycle costs, are higher than those for other designs.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £66 +P&P)