A focused guide to 15 peer-reviewed Building Simulation articles—organized by research question, connected to original DOI records, and supported by source-attributed figures.
The collection follows the path from sensing and forecasting to control, design and system operation. It highlights where AI models add value, where physical knowledge is embedded, and how uncertainty is handled.
Every paper has a stable page with verified bibliographic metadata, a concise problem–method–evidence summary, a representative figure and a direct DOI link.
The same paper can appear in more than one theme when it connects modeling, control or design.
Forecasting systems that combine ensembles, interpretability and specialized experts with design decisions.
Open topic synthesisAppliance- and HVAC-level load separation that exposes flexibility and demand-response potential.
Open topic synthesisOccupancy, AC behavior and stochastic disturbances translated into better operational control.
Open topic synthesisPhysical structure embedded in neural models to improve robustness, generalization and control reliability.
Open topic synthesisAI applied upstream to morphology, microclimate, layout generation and adaptive reuse.
Open topic synthesisEV data, day-ahead operation and uncertainty quantification for connected building energy systems.
Open topic synthesisRepresentative figures are excerpted from the supplied papers and linked to a complete source register.
Which AI models improve building-energy prediction while remaining interpretable?
How can appliance or HVAC demand be disaggregated without dense submetering?
How are occupancy forecasts and stochastic disturbances used in building control?
What does physics-informed learning add to robustness, extrapolation and reliability?
How can AI support microclimate prediction, layout generation and adaptive reuse?
How are EV data and design-load uncertainty handled in smart building systems?
Search titles, methods, authors, keywords or DOI records; then narrow the list by theme.
This paper connects predictive accuracy, model interpretation, and multi-objective design optimization in one workflow.
The study provides a hybrid deep-learning route for energy disaggregation and appliance-level load tracking.
The paper links unsupervised load disaggregation directly to operational flexibility assessment.
The study connects behavior prediction, multi-sensor fusion, and occupant-centered HVAC control.
The paper shows how urban morphology and wind-aware microclimate prediction can improve building-energy inputs.
The work applies a collaborative mixture-of-experts architecture to heterogeneous building-energy forecasting.
The paper integrates physics-based dynamics, data-driven disturbance prediction, and predictive control.
The study connects generative AI, energy prediction, and evolutionary optimization for energy-aware design.
The paper supports adaptive-reuse decisions with real-building validation and sensitivity analysis.
The work shifts physics-informed learning from loss-only constraints toward explicit physics-based information flow inside the network.
The study evaluates physical consistency not only for prediction but also for control, comfort, and grid interaction.
The paper combines continuous-time learning, physical structure, and explicit robustness analysis for small-data energy modeling.
The paper treats data quality as a first-class component of smart-building and EV load forecasting.
The work links building thermal flexibility, EV aggregation, market interaction, and distributionally robust optimization.
The paper applies graph representation learning to fast, geometry-aware uncertainty quantification at design stage.
The same 15 records are available as structured data and citation files. Each format uses the stable paper URLs and DOI identifiers shown on the site.