Smart Meters Time Series Clustering for Demand Response Applications in the Context of High Penetration of Renewable Energy Resources

Abstract

The variability in generation introduced in the electrical system by an increasing share of renewable technologies must be addressed by balancing mechanisms, demand response being a prominent one. In parallel, the massive introduction of smart meters allows for the use of high frequency energy use time series data to segment electricity customers according to their demand response potential. This paper proposes a smart meter time series clustering methodology based on a two-stage k-medoids clustering of normalized load-shape time series organized around the day divided into 48 time points. Time complexity is drastically reduced by first applying the k-medoids on each customer separately, and second on the total set of customer representatives. Further time complexity reduction is achieved using time series representation with low computational needs. Customer segmentation is undertaken with only four easy-to-interpret features: average energy use, energy–temperature correlation, entropy of the load-shape representative vector, and distance to wind generation patterns. This last feature is computed using the dynamic time warping distance between load and expected wind generation shape representative medoids. The two-stage clustering proves to be computationally effective, scalable and performant according to both internal validity metrics, based on average silhouette, and external validation, based on the ground truth embedded in customer surveys.

Segmentation using Customers Lifetime Value: Hybrid K-means Clustering and Analytic Hierarchy Process

“…Based on this research, the focus is on customer segmentation so that the marketing strategy implemented can maximise the use of electricity provided by the company. Previous research only focused on classifying customers based on patterns and types of electricity use, electricity demand, and the use of the K-Means grouping model [2], [9], [10], [12].…”

“…Previous Segmentation Studies Based on Electricity Consumption Data Table 1 presents previous studies on customer segmentation using electricity consumption data, including the exploration of variables in the segmentation analyses [2], [3], [6]. It can also be seen that K-Means Clustering has been a popular technique [8], [9]. McLoughlin et al 2015 [2] profiled electricity load using experimental data by installing 4,000 intelligent meters in several homes in Ireland.…”

“…They compared the models to discover the patterns of customers' daily electricity usage. Banales et al 2021 [9] estimated energy reserves using customer electricity requests. The study applied the K-medoid model and silhouette method to customers' half-day electricity usage considering wind energy as an alternative energy source.…”

“…After the number of sets is determined, the processing begins with randomly generated centroids and iteratively calculating new centroids to gather to the last group. The steps in the k-means model are described as follows [9].…”

Design and Implementation of a Smart Energy Meter Using a LoRa Network in Real Time

“…Nainar et al [42] created an SM for monitoring low-voltage networks, applying a measurement estimation algorithm. Bañales et al [43] proposed a clustering method based on two-stage K-medoids clustering with normalised load profiles organised in time series. Zhao et al [44] addressed the problem of improving the efficiency of malfunction detection in SMs using deep learning techniques.…”