Non-Intrusive Load Monitoring for Energy Consumption Disaggregation

Abstract

The smart grid offers a venue for reducing the disparity in demand and generation by demand response initiatives. The efficacy of demand response algorithms relies on identifying the active non-essential loads at consumer premises during peak hours. Hence, separating the electricity usage of a household into its individual appliance consumption is essential for facilitating demand response. Non-intrusive load monitoring (NILM) is the widely adopted methodology for the disaggregation of power consumption. This would consequently help the consumers to manage their energy usage. This paper has implemented and compared two deep learning architectures, CNN and Bi-GRU network for energy consumption disaggregation. Standard UKDALE dataset is used for the training and testing of these architectures. The complex nature of the Bi-GRU network identified appliances with sporadic activity nature whereas CNN performed better in appliances that exhibit periodicity.