Exhaustive search; does it have a role in explorative design?
conference contributionposted on 29.07.2016 by Jonathan Wright, Elli Nikolaidou, Christina Hopfe
Any type of content contributed to an academic conference, such as papers, presentations, lectures or proceedings.
Building performance simulation (BPS) is used routinely in design practice to evaluate the performance of candidate design solutions. However, two sources of uncertainty exist in the design process: in the selection of an optimum design solution; and in the predicted performance of the building (say, due to uncertain boundary conditions). These uncertainties can be evaluated and reduced through the use of an “explorative design” process, in which uncertainty quantification, multi-objective optimization, and sensitivity analysis are combined to provide information on the choice of robust and optimal design solutions. This paper investigates the use of an exhaustive search method to sample all combinations of design solutions and uncertain boundary conditions. The number of samples, and therefore the range of designs considered, are limited by the computation time of BPS. However, this paper concludes that design standards can be used to identify a viable range of design options, and that an exhaustive search applied to a limited design space provided enough information to identify and select robust design solutions. The paper also demonstrates the use of a new approach to identifying robust solutions that are guaranteed to remain optimal, regardless of the prevailing uncertainty in the boundary conditions.
This research is funded by the UK Engineering and Physical Sciences Research Council (EPSRC).
- Architecture, Building and Civil Engineering