Building Energy Prediction Using Artificial Neural Networks (LSTM)

Energy Consumption ◽

Energy Use ◽

Prediction Models ◽

Short Term Memory ◽

Significant Proportion ◽

Cooling Load ◽

Energy Prediction ◽

Abstract Modern buildings account for a significant proportion of global energy consumption worldwide. Therefore, accurate energy use forecast is necessary for energy management and conservation. With the advent of smart sensors, a large amount of accurate energy data is available. Also, with the advancements in data analytics and machine learning, there have been numerous studies on developing data-driven prediction models based on Artificial Neural Networks (ANNs). In this work a type of ANN called Large Short-Term Memory (LSTM) is used to predict the energy use and cooling load of an existing building. A university administrative building was chosen for its typical commercial environment. The network was trained with one year of data and was used to predict the energy consumption and cooling load of the following year. The mean absolute testing error for the energy consumption and the cooling load were 0.105 and 0.05. The percentage mean accuracy was found to be 92.8% and 96.1%. The process was applied to several other buildings in the university and similar results were obtained. This indicates the model can successfully predict the energy consumption and cooling load for the buildings studied. The further improvement and application of this technique for optimizing building performance are also explored.

Forecasting Energy Use in Buildings Using Artificial Neural Networks: A Review

Energies ◽

10.3390/en12173254 ◽

2019 ◽

Vol 12 (17) ◽

pp. 3254 ◽

Cited By ~ 30

Author(s):

Jason Runge ◽

Radu Zmeureanu

Keyword(s):

Neural Networks ◽

Energy Consumption ◽

Energy Use ◽

Building Energy ◽

Data Driven ◽

Past Century ◽

Energy Forecasting ◽

The Past ◽

During the past century, energy consumption and associated greenhouse gas emissions have increased drastically due to a wide variety of factors including both technological and population-based. Therefore, increasing our energy efficiency is of great importance in order to achieve overall sustainability. Forecasting the building energy consumption is important for a wide variety of applications including planning, management, optimization, and conservation. Data-driven models for energy forecasting have grown significantly within the past few decades due to their increased performance, robustness and ease of deployment. Amongst the many different types of models, artificial neural networks rank among the most popular data-driven approaches applied to date. This paper offers a review of the studies published since the year 2000 which have applied artificial neural networks for forecasting building energy use and demand, with a particular focus on reviewing the applications, data, forecasting models, and performance metrics used in model evaluations. Based on this review, existing research gaps are identified and presented. Finally, future research directions in the area of artificial neural networks for building energy forecasting are highlighted.

Bankruptcy Prediction Models: A Bet on Artificial Neural Networks

Global Journal For Research Analysis ◽

10.15373/22778160/february2014/16 ◽

2012 ◽

Vol 3 (2) ◽

pp. 48-50

Author(s):

Ana Isabel Velasco Fernández ◽

◽

Ricardo José Rejas Muslera ◽

Juan Padilla Fernández-Vega ◽

María Isabel Cepeda González

Keyword(s):

Neural Networks ◽

Prediction Models ◽

Bankruptcy Prediction ◽

Comprehensive comparison of artificial neural networks and long short-term memory networks for rainfall-runoff simulation

Physics and Chemistry of the Earth Parts A/B/C ◽

10.1016/j.pce.2021.103026 ◽

2021 ◽

pp. 103026

Author(s):

Ganquan Mao ◽

Meng Wang ◽

Junguo Liu ◽

Zifeng Wang ◽

Kai Wang ◽

...

Keyword(s):

Neural Networks ◽

Short Term Memory ◽

Rainfall Runoff ◽

Short Term ◽

Runoff Simulation ◽

Term Memory ◽

Comprehensive Comparison ◽

Long Short Term Memory ◽

Modelling and Prediction of Monthly Global Irradiation Using Different Prediction Models

Energies ◽

10.3390/en14082332 ◽

2021 ◽

Vol 14 (8) ◽

pp. 2332

Author(s):

Cecilia Martinez-Castillo ◽

Gonzalo Astray ◽

Juan Carlos Mejuto

Keyword(s):

Neural Networks ◽

Random Forests ◽

Prediction Models ◽

Meteorological Station ◽

Ann Model ◽

Global Irradiation ◽

Artificial Neural ◽

Input Variables ◽

Using Data

Different prediction models (multiple linear regression, vector support machines, artificial neural networks and random forests) are applied to model the monthly global irradiation (MGI) from different input variables (latitude, longitude and altitude of meteorological station, month, average temperatures, among others) of different areas of Galicia (Spain). The models were trained, validated and queried using data from three stations, and each best model was checked in two independent stations. The results obtained confirmed that the best methodology is the ANN model which presents the lowest RMSE value in the validation and querying phases 1226 kJ/(m2∙day) and 1136 kJ/(m2∙day), respectively, and predict conveniently for independent stations, 2013 kJ/(m2∙day) and 2094 kJ/(m2∙day), respectively. Given the good results obtained, it is convenient to continue with the design of artificial neural networks applied to the analysis of monthly global irradiation.

Over-parameterisation, a major obstacle to the use of artificial neural networks in hydrology?

Hydrology and Earth System Sciences ◽

10.5194/hess-7-693-2003 ◽

2003 ◽

Vol 7 (5) ◽

pp. 693-706 ◽

Cited By ~ 44

Author(s):

E. Gaume ◽

R. Gosset

Keyword(s):

Neural Networks ◽

Conceptual Model ◽

Linear Model ◽

Case Studies ◽

Stream Flow ◽

River Flow ◽

Prediction Models ◽

Rainfall Runoff ◽

Abstract. Recently Feed-Forward Artificial Neural Networks (FNN) have been gaining popularity for stream flow forecasting. However, despite the promising results presented in recent papers, their use is questionable. In theory, their “universal approximator‿ property guarantees that, if a sufficient number of neurons is selected, good performance of the models for interpolation purposes can be achieved. But the choice of a more complex model does not ensure a better prediction. Models with many parameters have a high capacity to fit the noise and the particularities of the calibration dataset, at the cost of diminishing their generalisation capacity. In support of the principle of model parsimony, a model selection method based on the validation performance of the models, "traditionally" used in the context of conceptual rainfall-runoff modelling, was adapted to the choice of a FFN structure. This method was applied to two different case studies: river flow prediction based on knowledge of upstream flows, and rainfall-runoff modelling. The predictive powers of the neural networks selected are compared to the results obtained with a linear model and a conceptual model (GR4j). In both case studies, the method leads to the selection of neural network structures with a limited number of neurons in the hidden layer (two or three). Moreover, the validation results of the selected FNN and of the linear model are very close. The conceptual model, specifically dedicated to rainfall-runoff modelling, appears to outperform the other two approaches. These conclusions, drawn on specific case studies using a particular evaluation method, add to the debate on the usefulness of Artificial Neural Networks in hydrology. Keywords: forecasting; stream-flow; rainfall-runoff; Artificial Neural Networks

Artificial Neural Networks, Sequence-to-Sequence LSTMs, and Exogenous Variables as Analytical Tools for NO2 (Air Pollution) Forecasting: A Case Study in the Bay of Algeciras (Spain)

Sensors ◽

10.3390/s21051770 ◽

2021 ◽

Vol 21 (5) ◽

pp. 1770

Author(s):

Javier González-Enrique ◽

Juan Jesús Ruiz-Aguilar ◽

José Antonio Moscoso-López ◽

Daniel Urda ◽

Lipika Deka ◽

...

Keyword(s):

Neural Networks ◽

Short Term Memory ◽

Prediction Method ◽

Monitoring Network ◽

Exogenous Variables ◽

Prediction Horizon ◽

Artificial Neural ◽

Input Variables ◽

Analytical Tools

This study aims to produce accurate predictions of the NO2 concentrations at a specific station of a monitoring network located in the Bay of Algeciras (Spain). Artificial neural networks (ANNs) and sequence-to-sequence long short-term memory networks (LSTMs) were used to create the forecasting models. Additionally, a new prediction method was proposed combining LSTMs using a rolling window scheme with a cross-validation procedure for time series (LSTM-CVT). Two different strategies were followed regarding the input variables: using NO2 from the station or employing NO2 and other pollutants data from any station of the network plus meteorological variables. The ANN and LSTM-CVT exogenous models used lagged datasets of different window sizes. Several feature ranking methods were used to select the top lagged variables and include them in the final exogenous datasets. Prediction horizons of t + 1, t + 4 and t + 8 were employed. The exogenous variables inclusion enhanced the model’s performance, especially for t + 4 (ρ ≈ 0.68 to ρ ≈ 0.74) and t + 8 (ρ ≈ 0.59 to ρ ≈ 0.66). The proposed LSTM-CVT method delivered promising results as the best performing models per prediction horizon employed this new methodology. Additionally, per each parameter combination, it obtained lower error values than ANNs in 85% of the cases.