An integrative approach for using multidisciplinary design optimization in AEC

The demand for energy and resource efficiency combined with cost efficiency calls for the application of optimization strategies in building design. These environmental and economic objectives require the integration of information from all involved disciplines and the consideration of the respective interdependencies. The technique of Multidisciplinary Design Optimization (MDO) as developed in aerospace technologies offers a tool for an approach of optimization in designing. However, MDO as developed in aerospace engineering mainly addresses technical design and strongly focuses on quantitative aspects. Consequently, for its application in building design, the development of new methods is required. • Quantitative and qualitative objectives. MDO in building design needs to consider the categories of quantitative as well as qualitative aspects in the objective function. The first category includes aspects that exist as quantities such as physical parameters, costs, or resource quantities. For these aspects, a direct consideration in objective and constraint functions is possible. The second category includes qualitative aspects such as function, appearance, indoor qualities, etc. that are of major importance for architectural design. The numerical nature of the MDO methods excludes the qualitative aspects from direct consideration. • Preference criteria, interactive optimization, and implicit constraints. For a part of the qualities, quantification by preference criteria is possible. An example is the definition of a good aspect ratio of elements in the façade. The setup of preference criteria needs the individual assessment of the situation by the designer. For the non-quantifiable aspects, an interactive process is required. In this process, the designer sets the characteristics of the model that optimization keeps constant as implicit constraints. • Component-oriented decomposition. The interactive procedure of design optimization requires an adequate structure of the building model. This component-oriented model needs to embed the mathematical description of objectives and constraints. This novel model structure aims at enabling the same immediate and interactive way of modeling of the optimization model as practiced in current CAD modelling. The decomposition, i.e. the translation of the design case to an optimization model, uses a component-oriented method instead of the discipline-oriented breakdown typical for MDO. An integrative model approach brings together the component-oriented structure common in building design with the decomposition of the optimization, i.e. the model definition by parameters, mathematical functions, and algorithmic procedures. • Extensions for a dynamic building model. The use of existing standards of modeling in building design as basis for the component-oriented decomposition facilitates the integration in CAD. An extension for optimization, presented with reference to the Industry Foundation Classes (IFC) as leading standard for building modeling, serves this purpose. The main change by the extensions is An integrative approach for using multidisciplinary design optimization in AEC