Abstract:As the global energy demand rises and climate change creates more challenges, optimizing the performance of non-residential buildings becomes essential. Traditional simulation-based optimization methods often fall short due to computational inefficiency and their time-consuming nature, limiting their practical application. This study introduces a new optimization framework that integrates Bayesian optimization, XGBoost algorithms, and multi-objective genetic algorithms (GA) to enhance building performance metrics—total energy (TE), indoor overheating degree (IOD), and predicted percentage dissatisfied (PPD)—for historical (2020), mid-future (2050), and future (2080) scenarios. The framework employs IOD as a key performance indicator (KPI) to optimize building design and operation. While traditional indices such as the predicted mean vote (PMV) and the thermal sensation vote (TSV) are widely used, they often fail to capture individual comfort variations and the dynamic nature of thermal conditions. IOD addresses these gaps by providing a comprehensive and objective measure of thermal discomfort, quantifying both the frequency and severity of overheating events. Alongside IOD, the energy use intensity (EUI) index is used to assess energy consumption per unit area, providing critical insights into energy efficiency. The integration of IOD with EUI and PPD enhances the overall assessment of building performance, creating a more precise and holistic framework. This combination ensures that energy efficiency, thermal comfort, and occupant well-being are optimized in tandem. By addressing a significant gap in existing methodologies, the current approach combines advanced optimization techniques with modern simulation tools such as EnergyPlus, resulting in a more efficient and accurate model to optimize building performance. This framework reduces computational time and enhances practical application. Utilizing SHAP (SHapley Additive Explanations) analysis, this research identified key design factors that influence performance metrics. Specifically, the window-to-wall ratio (WWR) impacts TE by increasing energy consumption through higher heat gain and cooling demand. Outdoor temperature (Tout) has a complex effect on TE depending on seasonal conditions, while indoor temperature (Tin) has a minor impact on TE. For PPD, Tout is a major negative factor, indicating that improved natural ventilation can reduce thermal discomfort, whereas higher Tin and larger open areas exacerbate it. Regarding IOD, both WWR and Tin significantly affect internal heat gains, with larger windows and higher indoor temperatures contributing to increased heat and reduced thermal comfort. Tout also has a positive impact on IOD, with its effect varying over time. This study demonstrates that as climate conditions evolve, the effects of WWR and open areas on TE become more pronounced, highlighting the need for effective management of building envelopes and HVAC systems.

Decision-Oriented Modeling of Thermal Dynamics Within Buildings

A Multiple Model Approach for Predictive Control of Indoor Thermal Environment with High Resolution

Modeling and Optimization of Complex Building Energy Systems with Deep Neural Networks

A Lifelong Meta-Learning Approach for Learning Deep Grey-box Representative Thermal Dynamics Models for Residential Buildings

Physics Informed Neural Networks for Control Oriented Thermal Modeling of Buildings

Ventilation and Temperature Control for Energy-efficient and Healthy Buildings: A Differentiable PDE Approach

Combining Physics-based and Data-driven Modeling for Building Energy Systems

Deep Transfer Learning for Thermal Dynamics Modeling in Smart Buildings

Ensemble machine learning for managing the required thermal energy from the architectural characteristics of residential buildings

Control-Oriented, Data-Driven Models of Thermal Dynamics

End-to-end deep meta modelling to calibrate and optimize energy consumption and comfort

Integrating Machine Learning and Genetic Algorithms to Optimize Building Energy and Thermal Efficiency Under Historical and Future Climate Scenarios

Energy efficiency in cooling systems: integrating machine learning and meta-heuristic algorithms for precise cooling load prediction

Optimising Building Energy and Comfort Predictions with Intelligent Computational Model

Neural differential equations for temperature control in buildings under demand response programs

Energy-Efficient Thermal Comfort Control in Smart Buildings via Deep Reinforcement Learning

Real-life Data-driven Model Predictive Control for Building Energy Systems Comparing Different Machine Learning Models

A Distributed ADMM-based Deep Learning Approach for Thermal Control in Multi-Zone Buildings under Demand Response Events

Model predictive control with adaptive machine-learning-based model for building energy efficiency and comfort optimization

Unlocking Energy Flexibility From Thermal Inertia of Buildings: A Robust Optimization Approach

Online Model Estimation for Predictive Thermal Control of Buildings