Development of surrogate models using artificial neural network for building shell energy labelling

Melo, A.P.; Cóstola, D Daniel; Lamberts, Roberto; Hensen, Jlm Jan

doi:10.1016/j.enpol.2014.02.001

Cited by 71 publications

(40 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…When real data are not available, energy simulations are carried out to examine the building samples [11,[21][22][23]. In order to extrapolate the outcomes to large stocks, statistical tools can be used, such as regression models [24,25] and meta-model techniques, e.g., artificial neural networks [26][27][28]. Notably, when bottom-up approaches are implemented, the considered sample has to be representative of the explored building stock, and therefore it should be carefully defined in terms of size and composition, depending on the characteristics of the stock [29].…”

Section: Aim and Originality Of The Studymentioning

confidence: 99%

Addressing Large-Scale Energy Retrofit of a Building Stock via Representative Building Samples: Public and Private Perspectives

et al. 2017

View full text Add to dashboard Cite

Scientific literature about energy retrofit focuses on single buildings, but the investigation of whole building stocks is particularly worthy because it can yield substantial energy, environmental and economic benefits. Hence, how to address large-scale energy retrofit of existing building stocks? The paper handles this issue by employing a methodology that provides a robust energy analysis of building categories. This is denoted as SLABE, "Simulation-based Large-scale uncertainty/sensitivity Analysis of Building Energy performance". It was presented by the same authors and is here enhanced to investigate a whole and heterogeneous building stock that includes various categories. Each category is represented via a Representative Building Sample (RBS), which is defined through Latin hypercube sampling and uncertainty analysis. Hence, optimal retrofit packages are found in function of building location, intended use and construction type. Two families of optimal solutions are achieved. The first one collects the most energy-efficient (and thus sustainable) solutions, among the ones that produce global cost savings, thereby addressing the public perspective. The second one collects cost-optimal solutions thereby addressing the private perspective. EnergyPlus is employed as a simulation tool and coupled with MATLAB ® for data analysis and processing. The methodology is applied to a significant share of the Italian public administration building stock, which includes several building categories depending on location, use destination and construction type. The outcomes show huge potential energy and economic savings, and could support a deep energy renovation of the Italian building stock.

show abstract

Section: Aim and Originality Of The Studymentioning

confidence: 99%

Addressing Large-Scale Energy Retrofit of a Building Stock via Representative Building Samples: Public and Private Perspectives

et al. 2017

View full text Add to dashboard Cite

show abstract

“…They are not programmed in the traditional way but they are trained using past history data representing the behaviour of a system [14]. ANNs can be defined as the learning, understanding and thinking ability of computers [10] that have been widely used for a range of applications in the area of energy modelling [11,[30][31][32][33][34][35][36][37]. Several studies published on predicting energy consumption, cooling load and heat load of buildings using ANN methods [23,[30][31][32][33].…”

Section: Introductionmentioning

confidence: 99%

“…The results showed that ANN model results fairly agree with the simulated data. Melo et al [36] presented an ANN model to improve the accuracy of surrogate models for building shell energy labelling purposes. Building properties with different areas, numbers of floor, conditioned areas and other characteristics were taken into account for the construction of the model.…”

Section: Introductionmentioning

confidence: 99%

“…Many authors [9,11,23,30,32,36,37] considered various input parameters for their models to estimate energy consumption and heat load. Hygh et al [9] used 27 building parameters including orientation, aspect ratio and shading projection while Wong et al [30] used external weather conditions (daily average dry-bulb temperature, daily average wet-bulb temperature, daily global solar radiation and daily average clearness index), building envelope designs (solar aperture, daylight aperture, overhang and side-fins projections) and day time (i.e.…”

Section: Introductionmentioning

confidence: 99%

“…Various authors [23,30,36,40] performed a comparison of ANN models with building dynamic simulation software. Neto and Fiorelli [23] presented a comparison of ANN with a physical simulation tool as EnergyPlus.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Comparative study of a building energy performance software (KEP-IYTE-ESS) and ANN-based building heat load estimation

Turhan

Kazanasmaz

Uygun

et al. 2014

Energy and Buildings

View full text Add to dashboard Cite

a b s t r a c tThe several parameters affect the heat load of a building; geometry, construction, layout, climate and the users. These parameters are complex and interrelated. Comprehensive models are needed to understand relationships among the parameters that can handle non-linearities. The aim of this study is to predict heat load of existing buildings benefiting from width/length ratio, wall overall heat transfer coefficient, area/volume ratio, total external surface area, total window area/total external surface area ratio by using artificial neural networks and compare the results with a building energy simulation tool called KEP-IYTE-ESS developed by Izmir Institute of Technology. A back propagation neural network algorithm has been preferred and both simulation tools were applied to 148 residential buildings selected from 3 municipalities of Izmir-Turkey. Under the given conditions, a good coherence was observed between artificial neural network and building energy simulation tool results with a mean absolute percentage error of 5.06% and successful prediction rate of 0.977. The advantages of ANN model over the energy simulation software are observed as the simplicity, the speed of calculation and learning from the limited data sets.

show abstract

Introducing Cost Models to Conceptual Tradespace Exploration

Baylot

Kelley

Richards

et al. 2018

INCOSE International Symp

View full text Add to dashboard Cite

Over 70% of lifecycle cost (LCC) for a system design is determined during the conceptual phase. Currently for a US Department of Defense (DoD) Capabilities‐Based Assessment (C‐BA) or an Analysis of Alternatives (AoA), the cost analysis is performed on a small set of chosen designs based on performance/attribute metrics. Ideally, it would be better to explore the cost versus performance trades on thousands, if not millions, of designs before the down‐selection to reveal the more affordable designs. A goal of the DoD Engineered Resilient Systems (ERS) Program is to create a capability that links cost and performance models for early concept exploration. ERS leverages advances in computer science and computing power to create a tradespace framework to couple performance and cost model results. Integrating cost analysis into early tradespace development aligns with the evolving DoD policy to formally include cost earlier in the acquisition timeline.

show abstract

Development of surrogate models using artificial neural network for building shell energy labelling

Cited by 71 publications

References 78 publications

Addressing Large-Scale Energy Retrofit of a Building Stock via Representative Building Samples: Public and Private Perspectives

Addressing Large-Scale Energy Retrofit of a Building Stock via Representative Building Samples: Public and Private Perspectives

Comparative study of a building energy performance software (KEP-IYTE-ESS) and ANN-based building heat load estimation

Introducing Cost Models to Conceptual Tradespace Exploration

Contact Info

Product

Resources

About