The design of heating, ventilation, and air conditioning systems based on building information modeling: A review from the perspective of automatic and intelligent methods

Heating, ventilation, and air conditioning (HVAC) systems constitute a significant portion of building energy usage and carbon emissions. The design quality directly affects the energy-saving and carbon-reduction effects of HVAC systems. A high-quality HVAC system design necessitates the use of automatic and intelligent methods, as traditional HVAC system design is experiencebased, repetitive, time-consuming, and difficult to achieve optimality. Building information modeling (BIM) offers a standardized data basis for the implementation of automatic and intelligent methods in HVAC system design practice. This study aims to assess the current status, advantages, and disadvantages of automatic data interaction between BIM and HVAC system design simulation software, intelligent optimization methods for HVAC system design, and automatic clash filtering and resolution methods for mechanical, electrical and plumbing (MEP) systems. To this end, we conducted a comprehensive search of relevant scientific databases and reviewed 105 articles. The primary results indicate that automatic data interaction methods predominantly consist of IFC-based and gbXML-based approaches. HVAC system design optimization problems commonly employ metaheuristic algorithms, analytic hierarchy process (AHP), and Monte Carlo simulation. Additionally, machine learning and hybrid approaches that integrate graph-theory-based methods and optimization methods are extensively utilized for clash filtering and resolution in MEP systems. The findings of this review offer engineers automatic and intelligent approaches for addressing design challenges in practice. They also offer insights into the development of automatic and intelligent BIM-based HVAC system design, which can be beneficial for researchers and practitioners.