Genetic Algorithms for Optimization of Building Envelopes and the Design and Control of HVAC Systems

TLDR

Building design seeks to reduce capital and operating costs while maintaining service, and genetic algorithms—flexible yet sometimes less efficient than problem‑specific methods—have been successfully applied to such optimization tasks. The paper reviews GA fundamentals for multi‑objective building design and surveys applications such as window placement, wall composition, building form generation, and HVAC system design, while outlining future integration with simulation and CAD. The authors review GA basics and apply them to multi‑objective optimization of building envelope components and HVAC system design, illustrating how GAs can determine window placement, wall composition, building form, and HVAC operation.

Abstract

Many design problems related to buildings involve minimizing capital and operating costs while providing acceptable service. Genetic algorithms (GAs) are an optimization method that has been applied to these problems. GAs are easily configured, an advantage that often compensates for a sacrifice in performance relative to optimization methods selected specifically for a given problem, and have been shown to give solutions where other methods cannot. This paper reviews the basics of GAs, emphasizing multi-objective optimization problems. It then presents several applications, including determining the size and placement of windows and the composition of building walls, the generation of building form, and the design and operation of HVAC systems. Future work is identified, notably interfaces between a GA and both simulation and CAD programs.