TL;DR
This paper introduces an intelligent, automated framework for optimizing building atrium design to enhance sustainability and occupant comfort. By integrating parametric modeling, multi-objective simulation (daylighting, energy, thermal comfort), and genetic algorithms, the approach significantly reduces design time and improves performance metrics like useful daylight illuminance (UDI) by up to 43.20%.
What Problem Does This Paper Address?
Designing effective building atriums is crucial for creating sustainable and comfortable indoor environments, influencing natural ventilation, daylighting, and thermal regulation. However, traditional design methods rely heavily on architect experience, making it difficult to explore the vast design space and optimize for multiple performance objectives simultaneously. Existing parametric optimization studies often focus on simplified, rectangular atrium prototypes and neglect the interplay between daylighting, energy consumption, and thermal comfort. This paper addresses the need for a generalized, automated framework that can optimize complex and irregular atrium geometries for a holistic set of environmental performance criteria.
Methods
The core of this research is a parametric optimization framework built on the Rhino and Grasshopper platforms. Key methodological components include:
- Geometry Mapping: A novel method to transform any atrium shape into a rectangular profile. This allows for a generalized parametric system applicable to diverse building designs, overcoming the limitation of previous studies focusing on ideal prototypes.
- Parametric Design System: Six key geometric parameters are defined to represent atrium characteristics: Well Index (WI), Shape Ratio (SR), Volume Ratio (VR), Position Index X (PX), Position Index Y (PY), and Inner Interface Window-to-Wall Ratio (IWWR). These parameters are used to generate design variants.
- Performance Simulation: The framework integrates multiple simulation engines:
- Radiance for Useful Daylight Illuminance (UDI).
- EnergyPlus for Energy Use Intensity (EUI).
- A thermal comfort model (based on Predicted Percentage Dissatisfied - PPD) to calculate Discomfort Time Percentage (DTP).
- Multi-Objective Optimization (MOO): A genetic algorithm (specifically, the Wallacei plugin) is employed to iteratively search for optimal solutions that balance the three objectives (maximizing UDI, minimizing EUI, and minimizing DTP).
Datasets
The study utilizes a case study of a real building, "Villa Reden," located in Katowice, Poland. This building serves as the benchmark for evaluating the proposed framework. No new datasets were introduced; rather, the framework was applied to an existing architectural design.
Key Insights and Ideas
- Holistic Optimization is Key: The research demonstrates that optimizing for daylighting, energy, and thermal comfort simultaneously is essential, as these objectives often have complex interdependencies. For instance, improving daylight can sometimes increase energy loads.
- Geometry Mapping Unlocks Generalizability: The geometry mapping technique is a significant contribution, enabling the framework to be applied to irregular and complex atrium shapes, which are common in real-world architectural practice, unlike many prior studies limited to simple rectangular forms.
- Parametric Representation Matters: The selection of six specific geometric parameters (WI, SR, VR, PX, PY, IWWR) effectively captures the essential characteristics of an atrium's design, allowing for a comprehensive exploration of the design space.
- Trade-offs Exist: The study highlights that achieving optimal performance across all three objectives simultaneously can be challenging. For example, in moderate climates like Poland, maximizing UDI might lead to increased EUI, and occupant comfort (DTP) is heavily influenced by HVAC systems.
- Location and Shape are Crucial: The analysis reveals that specific atrium locations (PX, PY) and shapes (SR) can significantly impact both daylighting and thermal comfort, suggesting that these factors are as important as size or window area.
Key Results
- Significant Performance Improvements: The case study in Poland showed substantial improvements:
- Useful Daylight Illuminance (UDI): 43.20% improvement.
- Energy Use Intensity (EUI): 15.52% improvement.
- Discomfort Time Percentage (DTP): 3.89% improvement.
- Efficiency Gains: The automated optimization process is highly efficient, with an average time of 36 seconds per solution, drastically reducing the time and human cost compared to traditional manual methods (estimated 20 minutes per simulation). The total optimization for 1500 solutions took approximately 15 hours.
- Mapping Validation: The geometry mapping method was validated, showing high correlation (R² values of 0.991 for UDI, 0.818 for wEUI, and 0.960 for sDTP) between the mapped and original models, confirming its reliability.
- Best Solution Characteristics: The identified "overall best solution" for the case study featured a specific set of parameters (e.g., VR=1.15, WI=0.80, IWWR=0.60, SR=1.3125, PX=-0.125, PY=0.025) that balanced the three objectives, achieving a UDI of 35.14%, wEUI of 128.12 kWh/m², and sDTP of 58.14%.
Limitations and Future Work
- Mapping Deviation: While the geometry mapping method is effective, it may introduce some simulation deviation. Further research is needed to refine this mapping method for even greater accuracy.
- Parameter Sensitivity: Some parameters might have a limited impact on specific performance objectives, potentially affecting optimization efficiency. Future work could incorporate sensitivity analysis to better tune parameter influence and accelerate optimization.
- Climate Specificity: The study acknowledges that the impact of atrium design on thermal comfort (DTP) can be less pronounced in moderate climates where HVAC systems play a dominant role. Future work could explore this further in different climate zones.
- Further Parametric Development: The authors suggest that the framework can be extended by defining new parametric methods to achieve more specific optimization functions and explore a wider range of design requirements.
Summary updated