Prediction of Energy Consumption for Variable Customer Portfolios Including Aleatoric Uncertainty Estimation

Using hourly energy consumption data recorded by smart meters, retailers can estimate the day-ahead energy consumption of their customer portfolio. Deep neural networks are especially suited for this task as a huge amount of historical consumption data is available from smart meter recordings to be used for model training. Probabilistic layers further enable the estimation of the uncertainty of the consumption forecasts. Here, we propose a method to calculate hourly day-ahead energy consumption forecasts which include an estimation of the aleatoric uncertainty. To consider the statistical properties of energy consumption values, the aleatoric uncertainty is modeled using lognormal distributions whose parameters are calculated by deep neural networks. As a result, predictions of the hourly day-ahead energy consumption of single customers are represented by random variables drawn from lognormal distributions obtained as output from the neural network. We further demonstrate, how these random variables corresponding to single customers can be aggregated to probabilistic forecasts of customer portfolios of arbitrary composition.

[1]  Dustin Tran,et al.  TensorFlow Distributions , 2017, ArXiv.

[2]  Yi Wang,et al.  Review of Smart Meter Data Analytics: Applications, Methodologies, and Challenges , 2018, IEEE Transactions on Smart Grid.

[3]  Willem Waegeman,et al.  Aleatoric and epistemic uncertainty in machine learning: an introduction to concepts and methods , 2019, Machine Learning.

[4]  Sung Wook Baik,et al.  A Novel CNN-GRU-Based Hybrid Approach for Short-Term Residential Load Forecasting , 2020, IEEE Access.

[5]  Ran Li,et al.  Deep Learning for Household Load Forecasting—A Novel Pooling Deep RNN , 2018, IEEE Transactions on Smart Grid.

[6]  Jingrui Xie.  Probabilistic electric load forecasting , 2016 .

[7]  R. Weron,et al.  Recent advances in electricity price forecasting: A review of probabilistic forecasting , 2016 .

[8]  Rob J. Hyndman,et al.  Forecasting Uncertainty in Electricity Smart Meter Data by Boosting Additive Quantile Regression , 2016, IEEE Transactions on Smart Grid.

[9]  Marilyn A. Brown,et al.  Machine learning approaches for estimating commercial building energy consumption , 2017 .

[10]  Ning Zhang,et al.  Probabilistic individual load forecasting using pinball loss guided LSTM , 2019, Applied Energy.

[11]  Joakim Widén,et al.  Review on probabilistic forecasting of photovoltaic power production and electricity consumption , 2018 .

[12]  Enrico Carpaneto,et al.  Probabilistic characterisation of the aggregated residential load patterns , 2008 .

[13]  Jun Hu,et al.  Short-Term Load Forecasting With Deep Residual Networks , 2018, IEEE Transactions on Smart Grid.

[14]  Nora El-Gohary,et al.  A review of data-driven building energy consumption prediction studies , 2018 .

[15]  Ilkka Norros,et al.  Practical lognormal framework for household energy consumption modeling , 2015 .

[16]  Olaf Enge-Rosenblatt,et al.  Energy Consumption Curves of 499 Customers from Spain , 2021 .

[17]  Saad Mekhilef,et al.  Accurate Prediction of Hourly Energy Consumption in a Residential Building Based on the Occupancy Rate Using Machine Learning Approaches , 2021, Applied Sciences.

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

[19]  Jianqiang Yi,et al.  Building Energy Consumption Prediction: An Extreme Deep Learning Approach , 2017 .

[20]  Daniel Kirschen,et al.  Combining Probabilistic Load Forecasts , 2018, IEEE Transactions on Smart Grid.

[21]  Joakim Widén,et al.  Characterizing probability density distributions for household electricity load profiles from high-resolution electricity use data , 2014 .

[22]  Yi Wang,et al.  Using Bayesian Deep Learning to Capture Uncertainty for Residential Net Load Forecasting , 2020, IEEE Transactions on Power Systems.

[23]  Li Lin,et al.  Smart Meters in Smart Grid: An Overview , 2013, 2013 IEEE Green Technologies Conference (GreenTech).

[24]  Matthew J. Eckelman,et al.  Predictive modeling for US commercial building energy use: A comparison of existing statistical and machine learning algorithms using CBECS microdata , 2018 .