Collaborative optimisation in building design with a Pareto-based genetic algorithm

Large-scale building design is a constantly evolving discipline. Design managers are consistently trying to identify means for producing a better product in a shorter period of time. Hence there is a need for design assistant tools that can help designers understand the big picture. It is becoming hard to improve the system performance of building design based merely on advances in individual disciplines. In other words, improvements in individual disciplines alone are not sufficient to affect the improvements in the whole system. To achieve higher quality, system-orientated, holistic, multidisciplinary approaches to building design are needed (NSF, 1996). For this reason, this research investigates the applicability of multidisciplinary disciplinary optimisation (MDO) methodology in building design. The MDO methods divide a single system into a group of smaller sub-systems and effectively manage interactions between sub-systems. In the context of building design, the single system refers to the whole building design, and sub-system could be each disciplinary design. Such approaches could reduce the time and cost associated with the multidisciplinary design cycle. This thesis describes the work of developing collaborative optimisation framework with a Pareto based genetic algorithm (COPGA).