Optimization through classical design of experiments (DOE): An investigation on the performance of different factorial designs for multi-objective optimization of complex systems

Abstract

This study aims to expand the understanding and practical applications of classical Design of Experiments (DOE) techniques by exploring their potential, using polynomial response surface modeling, to optimize complex systems and tackle multi-objective optimization problems. Such issues are common across many engineering branches, and various building engineering problems can benefit from a better understanding and application of these techniques. Through a simulation-based study involving over 350′000 simulations in EnergyPlus, the study systematically evaluates more than 150 different factorial designs, comparing their performance in multi-objective optimization using a double-skin façade system as a case study. The findings indicate that different experimental designs vary in their success at optimizing façade performance, with central-composite designs performing best overall. Taguchi designs, while less reliable, are effective in identifying optimal levels of categorical factors. The study recommends using a central composite design if resources allow. For scenarios with many continuous factors, a screening design should be used initially to eliminate insignificant factors, followed by a central composite design for final optimization. When dealing with both continuous and categorical factors, a Taguchi design should first be applied to handle all levels of categorical factors and represent continuous factors in a two-level format. After determining the optimal levels of categorical factors, a central composite design should be used for the final optimization stage. The guidelines developed from this investigation offer deeper insights into the applicability and effectiveness of classical DOE in designing complex systems, not only for advanced building envelope systems but also for a broader range of building and non-building engineering systems. • Classical DOEs can optimize complex envelope systems for efficient buildings. • Central-composite designs excel in optimizing double-skin façades. • Taguchi designs find optimal categorical factors but are less reliable. • Effective optimization needs key continuous and categorical factors identified. • More criteria in the objective function increase optimization challenges.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 27 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.