Robust Smart Meter Data Analytics Using Smoothed ALS and Dynamic Time Warping

Abstract

This paper presents a robust data-driven framework for clustering large-scale daily chronological load curves from smart meters, with a focus on the challenges encountered in practice. The first challenge is the low data quality issue due to bad and missing data, which has been a major obstacle for various in-depth analyses of smart meter data. A novel Smoothed Alternating Least Squares (SALS) approach is proposed to recover missing/bad smart meter data by taking advantage of their low-rank property. The second challenge is brought by different data reporting rates of smart meters. A Dynamic Time Warping (DTW)-based approach is proposed that is more efficient and eliminates the need for data interpolation or measurement downsampling. The proposed approach enables flexible data collection strategies and gateway locations to meet various smart grid performance requirements. The proposed framework is tested through experiments using real-world smart meter data.

Analyzing Long-Term and High Instantaneous Power Consumption of Buildings from Smart Meter Big Data with Deep Learning and Knowledge Graph Techniques

“…Furthermore, the time granularity of electricity consumption records, especially generated by different brands of smart meters, often exhibits significant variation [28,29]. For example, occasional instances of missing records may occur, leading to situations where only one or two power consumption records are available per 15 min, despite the intended frequency of consumption measurement being at the minute level [30].…”

“…Consequently, within a specific time interval, such as 15 min, the quantity of accurately recorded electricity usage records can fluctuate [30]. The literature often discusses the sampling frequency of smart meters, which can vary significantly, ranging from as fast as thousands of samples per second (expressed in kHz) to as long as two hours [28,29]. Instant electricity consumption data collected over a few minutes are usually sufficient for most analyses, aiding residents in monitoring and conserving electricity usage.…”

Denoising Autoencoder-Based Missing Value Imputation for Smart Meters

“…Although it is possible to achieve finegrained data with reliable communications through advanced smart metering technologies, there still exist missing values in databases. Such missing values occur due to unexpected device power off, communication failure, measuring error, or other unknown reasons [11], [12]. Therefore, missing values should be properly recovered during the preprocessing stage.…”