Machine Learning Based Integrated Feature Selection Approach for Improved Electricity Demand Forecasting in Decentralized Energy Systems

Abstract

Improved performance electricity demand forecast can provide decentralized energy system operators, aggregators, managers, and other stakeholders with essential information for energy resource scheduling, demand response management, and energy market participation. Most previous methodologies have focused on predicting the aggregate amount of electricity demand at national or regional scale and disregarded the electricity demand for small-scale decentralized energy systems (buildings, energy communities, microgrids, local energy internets, etc.), which are emerging in the smart grid context. Furthermore, few research groups have performed attribute selection before training predictive models. This paper proposes a machine learning (ML)-based integrated feature selection approach to obtain the most relevant and nonredundant predictors for accurate short-term electricity demand forecasting in distributed energy systems. In the proposed approach, one of the ML tools -binary genetic algorithm (BGA) is applied for the feature selection process and Gaussian process regression (GPR) is used for measuring the fitness score of the features. In order to validate the effectiveness of the proposed approach, it is applied to various building energy systems located in the Otaniemi area of Espoo, Finland. The findings are compared with those achieved by other feature selection techniques. The proposed approach enhances the quality and efficiency of the predictor selection, with minimal chosen predictors to achieve improved prediction accuracy. It outperforms the other evaluated feature selection methods. Besides, a feedforward artificial neural network (FFANN) model is implemented to evaluate the forecast performance of the selected predictor subset. The model is trained using two-year hourly dataset and tested with another one-year hourly dataset. The obtained results verify that the FFANN forecast model based on the BGA-GPR FS selected training feature subset has achieved an annual MAPE of 1.96%, which is a very acceptable and promising value for electricity demand forecasting in small-scale decentralized energy systems.INDEX TERMS Binary genetic algorithm, decentralized energy system, electricity demand forecasting, feature selection, feedforward artificial neural network, fitness evaluation measure, Gaussian process regression, machine learning, smart grid.

Short-Term Forecasting of Heat Demand of Buildings for Efficient and Optimal Energy Management Based on Integrated Machine Learning Models

“…The BGA-GPR-based FS is executed for each of the four buildings. The obtained FS results are given in Table III. 2,3,4,5,6,7,8,9,12,14,16,18,19,20 As shown in Table III, the number of features selected by the devised FS method is noticeably smaller than the size of the predictor space (number of candidate variables in Table I). That means there have been insignificant and repeated information by most of the features in the original predictor space.…”

“…Occupancy represents the number of heat energy users (people) in the building at each time interval. It has been found that building energy consumptions have a high correlation with their user (people) occupancies [19]. However, due to limitation of time and resources for direct measurement of occupancy at the buildings, we have used indirect representation of buildings occupancy using two additional variables, the holiday indicator and period of the day.…”

“…Feature selection is the process of finding a cluster of the most significant features (attributes, variables, or predictors) for use in forecasting model development. It enhances prediction accuracy, reduces model training time and overfitting [19]. The major premise when using a FS is that the initial dataset contains some attributes that are either repetitive or superfluous, and can thus be eradicated without persuading loss of relevant information.…”

Designing Sustainable Smart Cities: Cooperative Energy Management Systems and Applications

“…A day-ahead market was started in 1990, with the England & Wales Electricity Pool in the United Kingdom being one of the world's leading markets [70]; to trade in such a market, it was essential to grasp and forecast demand and to plan the operation of generators in accordance with demand. In particular, there have been many studies on the forecast of total demand on the electric power system for various forecasting time horizons [71][72][73][74] and for various applications and situations [75][76][77]. Many time-series statistical modeling methods and data-driven prediction methods have been proposed.…”

Multi-Objective Particle Swarm Optimization Algorithm for Multi-Step Electric Load Forecasting

“…Using the cluster analysis method to divide 12-month data into seasonal-based data, and considering multi-input and multi-output, Bedi and Toshniwal [21] used long short-term memory (LSTM) to forecast the long-term electric load. Eseye et al [22] first used the binary genetic algorithm (BGA) to select features, then applied gaussian process regression (GPR) to evaluate the fitness function of the feature, and finally employed feedforward artificial neural network (FFANN) to predict the electric load. The annual MAPE value of the proposed method can reach 1.96%.…”