Researchers at Kanzawa University in Japan have developed a next-generation building energy evaluation model that combines rule-based symbolic AI computing with VR technology. The model enables real-time visualisation and simultaneous evaluation of the energy-saving effects and indoor thermal comfort during the design stage of a zero-energy building (ZEB). According to the researchers, the approach will have a wide range of applications in the design of next-generation smart buildings.
In recent years, in the context of global decarbonisation, reducing energy consumption in buildings has become an important issue. In designing a ZEB, achieving both energy efficiency and occupant comfort is a crucial challenge in realising sustainable architecture. However, static simulation methods widely used up until now have difficulty in fully evaluating changes in heat load and indoor environment during the design stage. This has led to uncertainty in design decisions. In particular, task-ambience air conditioning (TAAC) systems, which separately control the workspaces of individual occupants and the entire room, have been reported to save energy during operation. However, evaluation methods weren’t readily available for use at the design stage.
Advertisement
In this study, the researchers aimed to develop a digital-twin-type evaluation model that can be used from the design stage to visualise and evaluate in advance the energy-saving effects and comfort of air conditioning.
The research group, led by Professor Teng of the College of Transdisciplinary Sciences for Innovation, at Kanazawa University, in collaboration with a scientist from Fushou University, China, developed a rule-based symbolic AI-computing-driven digital twin model, dubbed VEEM-ZEB, that can simultaneously evaluate energy consumption and indoor thermal comfort during the architectural design stage (ex-ante) of a ZEB using a TAAC system.
Advertisement
The model is characterised by its ability to explicitly separate the thermal loads of task air conditioning and ambience air conditioning, and to perform integrated thermal comfort and energy evaluations based on PMV/PPD indices. Furthermore, the visualisation environment integrated with VR shows energy consumption and comfort indices in real time, realising a design-support mechanism that allows designers to instantaneously check evaluation results while manipulating conditions.
In addition, based on standardised parameter settings, the model is able to systematically generate and analyse about 48,000 design and operating scenarios. As a result of sensitivity analysis taking into account differences in seasonal conditions, occupancy density and behavioural modes, as well as demonstration in office spaces, this method was able to reduce air conditioning energy consumption by 7–8.5 per cent with a stable average annual energy-saving effect of 7.62 per cent.
Advertisement
The novelty of this research lies in the fact that it enables TAAC performance evaluation, which previously relied on the operational stage, to be carried out at the design stage, and that it has established a three-layered digital twin evaluation platform that integrates a highly explainable and reproducible computing method using rule-based symbolic AI with an intuitive visualisation environment using VR. This makes it possible to quantitatively compare air conditioning methods and control strategies at an early design stage.
The results of the research make it possible to simultaneously visualise and evaluate the energy-saving effects and indoor comfort of air conditioning systems at the design stage before building completion. The model is expected to be introduced into architectural design practice as a decision-making support tool for rational ZEB design that achieves both comfort and energy efficiency.
According to the researchers, the technology can be applied to a wide range of architectural applications, including offices and public, educational and medical facilities, and will contribute to reducing energy consumption and utility costs through the spreading of efficient air conditioning designs. Furthermore, as a pre-assessment method integrating rule-based symbolic AI and digital twins, the technology is expected to contribute to the development of architectural environmental engineering and ZEB design research.
Advertisement
The research has been published in Sustainable Cities and Society.