Prediction of residential building energy consumption: A neural network approach

Energy ◽  
2016 ◽  
Vol 117 ◽  
pp. 84-92 ◽  
Author(s):  
M.A. Rafe Biswas ◽  
Melvin D. Robinson ◽  
Nelson Fumo
Geofluids ◽  
2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Xuenan Zhang ◽  
Jinxin Zhang ◽  
Jinhua Zhang ◽  
YuChuan Zhang

As the energy consumption of residential building takes a large part in the building energy consumption, it is important to promote energy efficiency in residential building for green development. In order to evaluate the energy consumption of residential building more effectively, this paper proposes a combined prediction model based on random forest and BP neural network (RF-BPNN). To verify the prediction effect of the RF-BPNN combined model, experiments were performed by using the energy efficiency data set in the UCI database, and the model was evaluated with five indicators: mean absolute error, root mean square deviation, mean absolute percentage error, correlation coefficient, and coincidence index. Compared with the random forest, BP neural network model, and other existing models, respectively, it is proven by the experimental results that the RF-BPNN model possesses higher prediction accuracy and better stability.


Energies ◽  
2021 ◽  
Vol 14 (16) ◽  
pp. 4805
Author(s):  
Shu Chen ◽  
Zhengen Ren ◽  
Zhi Tang ◽  
Xianrong Zhuo

Globally, buildings account for nearly 40% of the total primary energy consumption and are responsible for 20% of the total greenhouse gas emissions. Energy consumption in buildings is increasing with the increasing world population and improving standards of living. Current global warming conditions will inevitably impact building energy consumption. To address this issue, this report conducted a comprehensive study of the impact of climate change on residential building energy consumption. Using the methodology of morphing, the weather files were constructed based on the typical meteorological year (TMY) data and predicted data generated from eight typical global climate models (GCMs) for three representative concentration pathways (RCP2.6, RCP4.5, and RCP8.5) from 2020 to 2100. It was found that the most severe situation would occur in scenario RCP8.5, where the increase in temperature will reach 4.5 °C in eastern Australia from 2080–2099, which is 1 °C higher than that in other climate zones. With the construction of predicted weather files in 83 climate zones all across Australia, ten climate zones (cities)—ranging from heating-dominated to cooling-dominated regions—were selected as representative climate zones to illustrate the impact of climate change on heating and cooling energy consumption. The quantitative change in the energy requirements for space heating and cooling, along with the star rating, was simulated for two representative detached houses using the AccuRate software. It could be concluded that the RCP scenarios significantly affect the energy loads, which is consistent with changes in the ambient temperature. The heating load decreases for all climate zones, while the cooling load increases. Most regions in Australia will increase their energy consumption due to rising temperatures; however, the energy requirements of Adelaide and Perth would not change significantly, where the space heating and cooling loads are balanced due to decreasing heating and increasing cooling costs in most scenarios. The energy load in bigger houses will change more than that in smaller houses. Furthermore, Brisbane is the most sensitive region in terms of relative space energy changes, and Townsville appears to be the most sensitive area in terms of star rating change in this study. The impact of climate change on space building energy consumption in different climate zones should be considered in future design strategies due to the decades-long lifespans of Australian residential houses.


2015 ◽  
Vol 77 (15) ◽  
Author(s):  
Jibrin Hassan Suleiman ◽  
Saeed Balubaid ◽  
Nasiru Mohammed Zakari ◽  
Egba Ernest Ituma

Most of the developing countries experience rapid urbanization and population growth, Malaysia is among these countries as the population and the energy consumption in the country tremendously increased over the last few decades.  A major challenge is the rate of energy consumption in the country is tremendous going higher which is a threat as the country was listed 26th out of the 30 top greenhouse emitters in the world.  A survey was conducted on the ways occupants’ consumes energy in their residential buildings in relation to dwelling factors in the State of Johor Malaysia. Energy consumption of the residential owners was assessed using drop and pick self-administered questionnaire. The questionnaires were answered by each household heads. Air conditioning system, refrigeration system, kitchen appliances, bathroom and laundry appliances, lighting appliances as well as other home appliances was considered in the survey. Correlation analysis was used using Statistical Package for Social Sciences (SPSS) to analyze the results. The finding shows a positive relationship between dwelling factors.  r ≥ 0.3 and above between dwelling factors and residential building energy consumption. 


2006 ◽  
Vol 5 (2) ◽  
pp. 407-412 ◽  
Author(s):  
Xindong Wei ◽  
Ji Xuan ◽  
Jun Yin ◽  
Weijun Gao ◽  
Bill Batty ◽  
...  

Buildings ◽  
2019 ◽  
Vol 9 (9) ◽  
pp. 204 ◽  
Author(s):  
Yang ◽  
Tan ◽  
Santamouris ◽  
Lee

With the rising focus on building energy big data analysis, there lacks a framework for raw data preprocessing to answer the question of how to handle the missing data in the raw data set. This study presents a methodology and framework for building energy consumption raw data forecasting. A case building is used to forecast the energy consumption by using deep recurrent neural networks. Four different methodologies to impute missing data in the raw data set are compared and implemented. The question of sensitivity of gap size and available data percentage on the imputation accuracy was tested. The cleaned data were then used for building energy forecasting. While the existing studies explored only the use of small recurrent networks of 2 layers and less, the question of whether a deep network of more than 2 layers would be performing better for building energy consumption forecasting should be explored. In addition, the problem of overfitting has been cited as a significant problem in using deep networks. In this study, the deep recurrent neural network is then used to explore the use of deeper networks and their regularization in the context of an energy load forecasting task. The results show a mean absolute error of 2.1 can be achieved through the 2*32 gated neural network model. In applying regularization methods to overcome model overfitting, the study found that weights regularization did indeed delay the onset of overfitting.


Sign in / Sign up

Export Citation Format

Share Document