Definition of Residential Power Load Profiles Clusters Using Machine Learning and Spatial Analysis

Mario Flor,Ivan Contreras,Sergio Herraiz

doi:10.3390/en14206565

Abstract

This study presents a novel approach for discovering actionable knowledge and exploring data-based models from data recorded by household smart meters. The proposed framework is supported by a machine learning architecture based on the application of data mining methods and spatial analysis to extract temporal and spatial restricted clusters of characteristic monthly electricity load profiles. In addition, it uses these clusters to perform short-term load forecasting (1 week) using recurrent neural networks. The approach analyses a database with measurements of 1000 smart meters gathered during 4 years in Guayaquil, Ecuador. Results of the proposed methodology led us to obtain a precise and efficient stratification of typical consumption patterns and to extract neighbour information to improve the performance of residential energy consumption forecasting.

Highlights

Power Load Profiles Clusters UsingSmart meters (SMs) provide a granularity and precision that make it possible to overcome classical methods such as the definition of a single typical consumption curve for residential or commercial clients
Clustering can be used on a daily electricity demand time series to group identical profiles to reveal the most typical load profiles [1,2]
As more long short-term memory (LSTM) hidden layers were added, the network was able to infer more complex behaviour in our time series and increase the accuracy of the prediction, so two hidden layers that assumed hourly and daily behaviour were used in our model

Summary

Introduction

Smart meters (SMs) provide a granularity and precision that make it possible to overcome classical methods such as the definition of a single typical consumption curve for residential or commercial clients. This rich source of data for energy consumption analysis shows that conventional methods are unable to cope with such volume or speed. This situation, together with the increasing availability of more powerful computers, has promoted the use of intelligent techniques to study patterns in time series data.

Objectives

Methods

Results

Discussion

Conclusion