Search

This paper presents the application of multi-objective genetic algorithms for holistic building design that considers multiple criteria; building energy use, capital cost, daylight distribution and thermal indoor environment. The optimization focus is related to building envelope parameters. To obtain relevant feedback from multi-objective optimizations in early design stages, evaluation speed is a key concern. The paper presents a fast evaluation method fit for the early design stages. It uses a combination of two different quasi-steady-state methods for energy and indoor environment evaluations, a Radiance implementation for daylight simulations and a scripted algorithm for capital cost evaluations. The application of the method is developed around an integrated dynamic model which allows visual design feedback from all evaluations to be an integrated part of the design tool experience. It is concluded, that quasi-steady-state methods implemented as part of integrated dynamic models are fast and flexible enough to support building energy-, indoor environment- and cost-optimization the early design stages. [ABSTRACT FROM AUTHOR]

Abstract: This paper presents a simple building simulation tool for integrated daylight and thermal analysis. The tool is capable of importing the thermal and visual properties for different glazings and shading positions from the Window Information System (WIS) program. A coupled ray-tracing and radiosity methodology is used to derive the daylight levels for different sky conditions. Both detailed daylight distribution for a particular day and time and hourly discrete values on a yearly basis may be obtained. For an integrated simulation the hourly daylight levels are fed into an existing simple thermal simulation program capable of calculating energy demand and the indoor environment. Straightforward control systems for general and task lighting systems have been implemented together with a shading control strategy that adjusts the shading according to the indoor operative temperature, the risk of glare and the profile angle of the sun. The implemented daylight calculation method allows for shades from the window recess and overhang, and for distant shades blocking the sky vault. Comparisons with the ray-tracing program Radiance show that the accuracy of this approach is adequate for predicting the energy implications of photoresponsive lighting control. The amount of input is small, which makes the tool useful for integrated daylight optimisation in the early design process. [Copyright &y& Elsevier]