Energy Consumption Forecasting for Smart Meters Using Extreme Learning Machine Ensemble.

Area of Science:

• Energy Systems Engineering
• Artificial Intelligence
• Data Science

Background:

• Smart meters are crucial for monitoring energy consumption, aiding power generation planning and management.
• Accurate energy consumption forecasting enhances decision-making and improves demand predictability for energy providers.

Purpose of the Study:

• To develop and evaluate a data-driven ensemble model for one-step-ahead energy consumption forecasting.
• To combine statistical and artificial neural network models for improved forecasting accuracy.

Main Methods:

• An ensemble model was created by combining five forecasting models: linear autoregressive, radial basis function, multilayer perceptron, extreme learning machines, and echo state networks.
• Extreme learning machines were utilized as the combination model for their efficiency and generalization capabilities.
• The ensemble model was trained and tested using real smart meter energy consumption data.

Main Results:

• The proposed ensemble model demonstrated superior performance compared to individual models and other existing methods.
• The model's effectiveness was validated using five distinct performance metrics.
• The ensemble approach significantly improved the accuracy of one-step-ahead energy consumption forecasts.

Conclusions:

• The developed data-driven ensemble model offers a highly effective solution for energy consumption forecasting.
• This approach enhances the predictability of energy demand, supporting efficient power generation and grid management.
• The findings highlight the potential of ensemble methods, particularly those leveraging extreme learning machines, in energy analytics.