Deep learning versus traditional machine learning methods for aggregated energy demand prediction

In this paper the more advanced, in comparison with traditional machine learning approaches, deep learning methods are explored with the purpose of accurately predicting the aggregated energy consumption. Despite the fact that a wide range of machine learning methods have been applied to probabilistic energy prediction, the deep learning ones certainly represent the state-of-the-art artificial intelligence methods with remarkable success in a spectrum of practical applications. In particular, the use of Multi Layer Perceptrons, recently enhanced with deep learning capabilities, is proposed. Furthermore, its performance is compared with the most commonly used machine learning methods, such as Support Vector Machines, Gaussian Processes, Regression Trees, Ensemble Boosting and Linear Regression. The analysis of the day-ahead energy prediction demonstrates that different prediction methods present significantly different levels of accuracy in the case of a challenging dataset that comprises an interesting mix of consumers, wind and solar generation. The results show that Multi Layer Perceptrons outperform all the eight methods used as a benchmark in this study.

[1]  Madeleine Gibescu,et al.  Demand forecasting at low aggregation levels using Factored Conditional Restricted Boltzmann Machine , 2016, 2016 Power Systems Computation Conference (PSCC).

[2]  Moncef Krarti.  Energy Audit of Building Systems : An Engineering Approach, Second Edition , 2016 .

[3]  Zhang Chao,et al.  Short-term prediction of wind power based on deep Long Short-Term Memory , 2016, 2016 IEEE PES Asia-Pacific Power and Energy Engineering Conference (APPEEC).

[4]  Sylvain Robert,et al.  State of the art in building modelling and energy performances prediction: A review , 2013 .

[5]  J. G. Slootweg,et al.  Improving Short-term load forecasting for a local energy storage system , 2012, 2012 47th International Universities Power Engineering Conference (UPEC).

[6]  Christopher K. I. Williams,et al.  Gaussian Processes for Machine Learning (Adaptive Computation and Machine Learning) , 2005 .

[7]  Léon Bottou,et al.  The Tradeoffs of Large Scale Learning , 2007, NIPS.

[8]  Hongseok Kim,et al.  Deep Neural Network Based Demand Side Short Term Load Forecasting , 2016 .

[9]  V. Ismet Ugursal,et al.  Comparison of neural network, conditional demand analysis, and engineering approaches for modeling end-use energy consumption in the residential sector , 2008 .

[10]  Dietmar Bruckner,et al.  A hidden Markov model based procedure for identifying household electric loads , 2011, IECON 2011 - 37th Annual Conference of the IEEE Industrial Electronics Society.

[11]  Frédéric Magoulès,et al.  A review on the prediction of building energy consumption , 2012 .

[12]  Madeleine Gibescu,et al.  Unsupervised energy prediction in a Smart Grid context using reinforcement cross-building transfer learning , 2016 .

[13]  Andrew L. Maas.  Rectifier Nonlinearities Improve Neural Network Acoustic Models , 2013 .

[14]  Daniel L. Marino,et al.  Building energy load forecasting using Deep Neural Networks , 2016, IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society.

[15]  J. Orbach.  Principles of Neurodynamics. Perceptrons and the Theory of Brain Mechanisms. , 1962 .

[16]  Tomah Sogabe,et al.  Optimization of decentralized renewable energy system by weather forecasting and deep machine learning techniques , 2016, 2016 IEEE Innovative Smart Grid Technologies - Asia (ISGT-Asia).

[17]  Milos Manic,et al.  Intelligent Buildings of the Future: Cyberaware, Deep Learning Powered, and Human Interacting , 2016, IEEE Industrial Electronics Magazine.

[18]  Anastasios I. Dounis,et al.  Artificial intelligence for energy conservation in buildings , 2010 .

[19]  P. H. Nguyen,et al.  Building Energy Management Systems — Optimization of comfort and energy use , 2013, 2013 48th International Universities' Power Engineering Conference (UPEC).

[20]  Geoffrey E. Hinton,et al.  Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.

[21]  Tony N.T. Lam,et al.  Artificial neural networks for energy analysis of office buildings with daylighting , 2010 .

[22]  G. Lewicki,et al.  Approximation by Superpositions of a Sigmoidal Function , 2003 .

[23]  Wil L. Kling,et al.  Comparison of machine learning methods for estimating energy consumption in buildings , 2014, 2014 International Conference on Probabilistic Methods Applied to Power Systems (PMAPS).

[24]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[25]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[26]  Matthew D. Zeiler.  ADADELTA: An Adaptive Learning Rate Method , 2012, ArXiv.

[27]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[28]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[29]  Soteris A. Kalogirou,et al.  Artificial neural networks in energy applications in buildings , 2006 .

[30]  Yunchao Wei,et al.  Deep Learning with S-Shaped Rectified Linear Activation Units , 2015, AAAI.

[31]  Madeleine Gibescu,et al.  Deep learning for estimating building energy consumption , 2016 .

[32]  Zongxiang Lu,et al.  Probabilistic short-term wind power forecasting based on deep neural networks , 2016, 2016 International Conference on Probabilistic Methods Applied to Power Systems (PMAPS).

[33]  L. Baum,et al.  Statistical Inference for Probabilistic Functions of Finite State Markov Chains , 1966 .