Electricity demand forecasting in industrial and residential facilities using ensemble machine learning

Abstract

This article presents electricity demand forecasting models for industrial and residential facilities, developed using ensemble machine learning strategies. Short term electricity demand forecasting is beneficial for both consumers and suppliers, as it allows improving energy efficiency policies and the rational use of resources. Computational intelligence models are developed for day-ahead electricity demand forecasting. An ensemble strategy is applied to build the day-ahead forecasting model based on several one-hour models. Three steps of data preprocessing are carried out, including treating missing values, removing outliers, and standardization. Feature extraction is performed to reduce overfitting, reducing the training time and improving the accuracy. The best model is optimized using grid search strategies on hyperparameter space. Then, an ensemble of 24 instances is generated to build the complete day-ahead forecasting model. Considering the computational complexity of the applied techniques, they are developed and evaluated on the National Supercomputing Center (Cluster-UY), Uruguay. Three different real data sets are used for evaluation: an industrial park in Burgos (Spain), the total electricity demand for Uruguay, and demand from a distribution substation in Montevideo (Uruguay). Standard performance metrics are applied to evaluate the proposed models. The main results indicate that the best day ahead model based on ExtraTreesRegressor has a mean absolute percentage error of 2:55% on industrial data, 5:17% on total consumption data and 9:09% on substation data. 

A Thermal Discomfort Index for Demand Response Control in Residential Water Heaters

“…This article extends our previous conference article "Demand response control in electric waterheaters: evaluation of impact on thermal comfort" [7], presented at III Iberoamerican Congress on Smart Cities. The main contributions beyond those of the previous conference article include: (i) an extension of the developed temperature model for electric water heaters, only measuring the electrical state of the device; (ii) an improved procedure to predict water utilization by applying data analysis to real electricity consumption data from the ECD-UY dataset; (iii) the definition of an index to approximate the discomfort associated with an active demand management interruption of the water heater; and (iv) an extended evaluation of the proposed methodology for several cases of active demand management interruption of water heaters over different realistic scenarios.…”

“…The split that computes the best result is then used to split the considered node of the tree [32]. ExtraTrees have proven to be an accurate predictor for electricity-related problems (e.g., for demand forecasting in industrial and residential facilities [7]).…”

“…Finally, after computing S uses (w) for each realization, TDI is computed as the empirical expected value defined in Equation (7).…”

CITIES: Ibero-American Research Network for Sustainable, Efficient, and Integrated Smart Cities

“…Some of the most relevant researches and published articles include the energy efficiency studies via computational intelligence for electricity demand forecasting in industrial and residential sectors [9]; IoT developments, such as the proposal of smart bus stops for increasing social inclusiveness and quality of life of elders [10]; electric mobility, such as a power supply solution for electric trains [11]; waste management in modern smart cities [12]; pollution and air quality, such as assessing the effectiveness of low emissions zones [13]; and sustainable mobility in public transportation [14].…”

LIMREF: Local Interpretable Model Agnostic Rule-Based Explanations for Forecasting, with an Application to Electricity Smart Meter Data

“…In the forecasting literature there are plenty of works on accurate electricity demand forecasting (Porteiro, Hernández-Callejo, and Nesmachnow 2020;Blum and Riedmiller 2013;Taylor 2003;Bedi and Toshniwal 2018;Al-Musaylh et al 2018). The traditional energy demand forecasting methods are state-of-the-art univariate forecasting approaches such as ARMA, Trend Seasonal (TBATS) models and ARIMA models.…”