Abstract
Demand Response Programs (DRPs) have emerged as effective tools to enhance grid stability, reduce peak demand, and promote energy efficiency. However, targeting the right consumers for DRPs remains a challenge, as different household’s exhibit varying consumption patterns. This research aims to develop a method for identifying optimal target groups for DRPs using clustering techniques, specifically K-means clustering, in combination with load profile analysis. The dataset used in this study is sourced from the National Renewable Energy Laboratory (NREL), based on the 2022 Residential Energy Consumption Survey (RECS), which includes power usage data for individual households recorded at 10-minute intervals. By analysing household electricity usage data and applying K-means clustering, consumers are grouped into distinct clusters based on their load profiles. The optimal number of clusters is determined using the Elbow method, which helps identify the most representative consumption patterns. The suitability of these clusters for various DRP strategies, such as time-of-use pricing or direct load control, is then evaluated. The findings demonstrate how clustering techniques can segment consumers effectively, allowing for more targeted and efficient DRP implementation. This approach optimizes participation in DRPs and contributes to reducing grid stress during peak demand periods and improving energy efficiency on a broader scale.
Keywords: Clustering, Demand Response, k-means, Load Profiling, Smart Grid, Time Series.