Item Infomation
| Title: | A New Interoperability Framework for Data-Driven Building Performance Simulation |
| Authors: | Malkawi, Ali Samuelson, Holly Protopapas, Pavlos |
| Description: | Machine learning (ML) and deep learning (DL) have become more prominent in the building, architecture, and construction industries. One area ideally suited to exploit this powerful new technology is building performance simulation (BPS) for sustainable building design. Physics-based models have traditionally been used to estimate the energy flow, air movement, and heat balance of buildings. The algorithms behind physics-based models, however, involve solving complex differential equations that require many assumptions, significant computational power, and a considerable amount of time to output predictions. With the advent of DL, which can handle large amounts of computation in a short period of time, data-driven models for predicting the physical properties of buildings are becoming increasingly popular due to their simplicity and efficiency. As such, artificial neural networks (ANNs) with measured or simulated data for environmental analysis are likely to be a more feasible option for designers during the early design phase. To train ANN models, 3D data is an asset to computer vision because they provide rich information about the geometry and the related environment. Depending on the 3D data representation considered, different challenges may emerge when using trained ANN models. Hence, an interoperability framework is required for converting building geometries and environment-related information into relevant 3D matrices for model training and utilization. However, to date there has been no research on this topic in the BPS field; thus, this research proposes a new data interoperability framework for ANN models with 3D buildings serving as inputs. The framework has been subjected to a trial investigation using several ANN modeling studies on radiation and airflow simulation. The result is a comprehensive process map that includes the BPS requirement for ANN modeling, related subprocesses (i.e., building geometry and environmental levels), specific rules and methods for modeling, and processing of input and output data. To accomplish this, data exchangers for the ANN models, geometry representation tool (GRT), and BIM specification tool (BST) were introduced and developed as computational tools. The comprehensive framework has been validated using the developed case studies, demonstrating its applicability for different Computer-aided design tools (i.e., Rhinoceros and Revit) and ANN models (i.e., radiation and airflow) and illustrating the future capacity of integrated ANNs to serve as a tool for use in BPS and early-stage modeling. |
| URI: | http://lib.yhn.edu.vn/handle/YHN/329 |
| Other Identifiers: | Han, Jung Min. 2022. A New Interoperability Framework for Data-Driven Building Performance Simulation. Doctoral dissertation, Harvard Graduate School of Design. 29210611 https://nrs.harvard.edu/URN-3:HUL.INSTREPOS:37372963 0000-0003-1429-6044 |
| Appears in Collections | Tài liệu ngoại văn |
ABSTRACTS VIEWS
11
VIEWS & DOWNLOAD
36
Files in This Item: