MULTI-CRITERIA OPTIMISATION OF PANELLED BUILDING ENVELOPES DURING CONCEPTUAL STAGES OF DESIGN USING EVOLUTIONARY ALGORITHMS

Abstract

The main focus of this dissertation is to investigate whether Evolutionary Algorithms (EA) can be used as a tool in the early phases of design for multi-criteria optimisation of panelled building envelopes. Buildings with panelled envelopes may be optimised in order create an appropriate relationship between the environmental filtering properties of the envelope and the layout of functional activities inside the building. However, only optimising the panel configuration is not sufficient. It may be possible to find better solutions by optimising other aspects of the design in tandem with the panel configuration. In particular, two other aspects that may be considered are the geometry of the external envelope and the internal layout of functional activities. For the purpose of this dissertation, two demonstrations are created. In the first demonstration, panel configuration and the geometry of the building envelope are optimised. In the second demonstration, panel configuration and layout of internal space arrangement are optimised. Both demonstrations use the same design scenario, which is based on a scenario presented in (Shea et al. 2006). In the research conducted by Shea et al. (2006), the design was optimised in order to achieve certain lighting requirements and constraints. However, in their research, the optimisation process only allowed panel configuration to be optimised. The demonstrations presented in this research will show that optimising panel configuration in tandem with other design aspects can result in further improvements in performance.