Artificial Neural Networks for parametric daylight design

Lorenz, Clara-Larissa; Spaeth, Achim Benjamin; Souza, Clarice Bleil De; Packianather, Michael

doi:10.1080/00038628.2019.1700901

Cited by 17 publications

(10 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Studies in the first category provide an in-depth description of adopting technologies for benefitting only specific tasks. These studies include acquiring highresolution geometry data of existing site conditions with 3D LiDAR scanning (Shih et al, 2019(Shih et al, , 2020, automatic generation and evaluation of a large number of design options using simulations and algorithms (Haymaker J et al, 2018;Lin & Gerber, 2014;Lorenz et al, 2020), improving design constructions with digital fabrication methods (Melenbrink et al, 2020;Wagner et al, 2020) and better understanding of built designs with data analytics (Fan et al, 2021;V E et al, 2021). They do not provide an overview of the benefits to a design project and a design practice.…”

Section: Method: Proposed Computation In Design (C-in-d) Frameworkmentioning

confidence: 99%

A framework for considering the use of computational design technologies in the built environment design process

Chen¹,

Janssen²,

Aviv³

et al. 2022

ITcon

View full text Add to dashboard Cite

This research proposes a Computation in Design (C-in-D) framework for design practices to consider the adoption of computational technologies in their design process. Examples of computational design technologies include but are not limited to computational algorithms, the Internet of Things (IoT), reality capture and digital fabrications. We develop the framework by categorizing design projects based on their physical scales, defining the work stages in a design project, and decomposing the design process into tasks and data flows. The computational design technologies can then be assessed by mapping its usage onto these data flows. The framework provides a basic structure for practices to customize and systematically assess the impacts of using computational design technologies in their design process. We demonstrate the C-in-D framework in three case studies, a sculpture design, an interior retrofit, and a form-based code development. The demonstration shows that web-based interactive parametric modeling and reality capture technologies can improve collaboration between the artist and modeler in the sculpture design project. IoT and optimization algorithms can improve the daylighting performance of the interior retrofit, and the use of Geographic Information System and reality capture technology can improve site analysis and visioning of the form-based code development process. The framework is a valuable tool for facilitating the adoption of new design technologies in practice.

show abstract

Section: Method: Proposed Computation In Design (C-in-d) Frameworkmentioning

confidence: 99%

A framework for considering the use of computational design technologies in the built environment design process

Chen¹,

Janssen²,

Aviv³

et al. 2022

ITcon

View full text Add to dashboard Cite

show abstract

“…This enabled all possible instances in the parameterized design solution space to be examined. The total simulation time was reduced by 65 % (Lorenz et al, 2020).…”

Section: Structure and Processmentioning

confidence: 99%

“…Aspect Method Description and conclusion Relevance (Lorenz et al, 2020) Visual comfort, sDA Grasshopper, fully connected artificial neural network, Levenberg-Marquardt Lorenz et al found great time saving potential when using ANNs for daylight simulations. Due to the achievable time-savings of 65%, ANNs offer a possibility to readapt the brute-force approach into the design process.…”

Section: Appendix 1 | Literature Matrixmentioning

confidence: 99%

“…Time series forecasting are an active research area, which have received a considerable amount of attention in the literature (Rozenberg et al, 2012). In parametric design environments, the use of ANNs promises greater feasibility than simulations for exploring the performance of different solutions, due to a reduction in overall computation time (Lorenz et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting annual illuminance and operative temperature in residential buildings using artificial neural networks

Kristiansen

Jamil

Hameed

et al. 2022

Building and Environment

View full text Add to dashboard Cite

“…On a smaller scale, there are studies in which daylight harvesting has been incorporated [25][26][27]. Regarding daylight, there are applications regarding operation of shading devices [28], of lighting system [29] and prediction of daylight sufficiency [30] based on illuminance simulations. Several techniques which make use of artificial intelligence (AI) have been employed for the metamodeling of energy performance, and many of these involve the use of artificial neural networks (ANNs) [31,32].…”

Section: Introductionmentioning

confidence: 99%

Metamodeling of the Energy Consumption of Buildings with Daylight Harvesting – Application of Artificial Neural Networks Sensitive to Orientation

Fonseca

Sc²,

Pereira

2021

Journal of Daylighting

View full text Add to dashboard Cite

Daylight harvesting is a well-known strategy to address building energy efficiency. However, few simplified tools can evaluate its dual impact on lighting and air conditioning energy consumption. Artificial neural networks (ANNs) have been used as metamodels to predict energy consumption with high precision, few input parameters and instant response. However, this approach still lacks the potential to estimate consumption when there is daylight harvesting, at the ambient level, where the effect of orientation can be noted. This study investigates this potential, in order to evaluate the applicability of ANNs as a tool to aid the architectonic design. The ANNs were approached as metamodels trained based on EnergyPlus thermo-energetic simulations. The network configuration focused on determining its simplest feasible form. The input parameters adopted as the main variables of the building envelope were as follows: orientation, window-to-wall ratio and visible transmission. The effects of the encoding of orientation as a network input parameter, the number of examples of each variable for network training and changing the parameters used for the training were evaluated. The networks predicted the individualized consumption according to the end use with errors below 5%, indicating their potential to be applied as a simplified tool to support the design process, considering the elementary variables of the building envelope. The discussion of results focused on guidelines and challenges to achieve this purpose when contemplating the broadening of the metamodel scope.

show abstract

Artificial Neural Networks for parametric daylight design

Abstract: 2019. Artificial neural networks for parametric daylight design. Architectural Science Review 63 (2) , pp.

Cited by 17 publications

References 23 publications

A framework for considering the use of computational design technologies in the built environment design process

A framework for considering the use of computational design technologies in the built environment design process

Predicting annual illuminance and operative temperature in residential buildings using artificial neural networks

Metamodeling of the Energy Consumption of Buildings with Daylight Harvesting – Application of Artificial Neural Networks Sensitive to Orientation

Contact Info

Product

Resources

About