A Hybrid Digital Twin and Evolutionary Model for Green Building Automation

Abstract

The strife for energy efficiency, along with enhancing comfort, has led to the usage of innovative technologies and
sustainable development in construction and architecture. One of the most advanced types is digital twin (DT)
technology, which provides a virtual dynamic representation of real buildings with current performance and predictive
real-time simulation and analysis. Also, robust optimization techniques and evolutionary algorithms (EAs) allow for
structural, multi-objective control systems like those in green buildings. Even though there are advancements in
sophisticated building and automation control systems (BMS), numerous automation strategies are rigid and deeprooted
in outdated models that don't adapt to automation or dynamic condition changes such as variability in
occupancy, weather conditions, and operational deviations. This paper proposes a new system model that mitigates
those outlined issues and these inefficiencies by combining a digital twin with an evolutionary optimizing engine. The
digital twin simulates building performance while synchronizing with real-time sensor data; simultaneously, the
evolutionary model develops optimal control strategies for HVAC, lighting, and ventilation control systems under
changing conditions. A case study of a medium-sized commercial building was used, with simulation outcomes
showing a 15-25% reduction in energy usage while demonstrating PMV-enhanced thermal comfort (in the range of
±0.5) and responsiveness by the adaptive system. This has resulted in energy-resilient and efficient operations and
automation of building systems.