Non-intrusive Load Monitoring Research Research Articles

A versatile, low-cost monitoring device suitable for non-intrusive load monitoring research purposes

Energy systems monitoring is a key factor in reducing global energy consumption. Although intrusive approaches can be adopted, Non-Intrusive Load Monitoring (NILM), which targets estimation of individual-appliance consumption from aggregated electricity measurements, is the main choice for retrofitting the plethora of non-smart electrical installations. Contemporary NILM requirements include high-frequency power-line sampling, efficient disaggregation algorithms that can, potentially, run locally, on the monitoring device, to offload the remote fog/cloud infrastructure, and various open datasets to allow the development of such algorithms. Towards all these directions, in this paper, a versatile monitoring device to support NILM research is presented. It offers high sampling frequency, sufficiency of local resources, and has been designed to be used in different environments (residential, commercial, industrial). Actually, it was deployed in all these environments, with multiple copies of it being used in two different industrial facilities for over two years. The developed device’s cost has been kept low, its accuracy has been experimentally verified to be sufficiently high, while its key hardware characteristics, in the context of NILM research, compare favorably with those of other devices in the literature. Along with the device, a residential, high-frequency dataset is also presented and made publicly available.

Evaluation of regression models and Bayes-Ensemble Regressor technique for non-intrusive load monitoring

Non-Intrusive Load Monitoring (NILM) is referred to as the task of decomposing the aggregated power load of a residential or commercial building into appliance-level consumption without the installation of dedicated smart meters. NILM, mostly considered as a supervised learning problem, is crucial for energy monitoring and management as this approach can detect load malfunction and prevent wastage of energy consumption. For most of regression based NILM research, not much work has been done in analyzing data against temporal parameters such as day of the week, week of the month, week of the year, and various quarters in a year. Additionally, various NILM research works focus on finding the best regression model for predicting appliance consumption but discards the fact that appliance usage varies on factors such as seasons, different working hours, and weekends. This paper aims to evaluate some regression algorithms used towards NILM research based on 8 different training and testing Methods which according to our knowledge covered major factors that affect the appliance usage. The dataset used for the evaluation of the regression models, were collected from a research lab at Grenoble INP, in Grenoble, France. The evaluation results show that instead of one algorithm, individual regression algorithms generated favorable outcomes for various appliances based on different train-test Methods. Furthermore, a novel Bayesian optimized Ensemble regressor model for predicting individual appliance consumption from aggregated load data is also proposed. Instead of just using the aggregated power information, the proposed model also uses temporal information of the dataset to estimate accurate consumption output of individual appliances. Extensive simulation results corroborate the merits of the proposed approach, which outperforms the benchmarking Methods. To ensure reproducibility of the results by the research community and a potential future improvement of the framework by other researchers, the complete source code is provided in the following repository: https://github.com/Mohammad-Kaosain-Akbar/NILM-Ensemble-Bayesian-Optimization

Performance-Aware NILM Model Optimization for Edge Deployment

Non-Intrusive Load Monitoring (NILM) describes the extraction of the individual consumption pattern of a domestic appliance from the aggregated household consumption. Nowadays, the NILM research focus is shifted towards practical NILM applications, such as edge deployment, to accelerate the transition towards a greener energy future. NILM applications at the edge eliminate privacy concerns and data transmission-related problems. However, edge resource restrictions pose additional challenges to NILM. NILM approaches are usually not designed to run on edge devices with limited computational capacity, and therefore model optimization is required for better resource management. Recent works have started investigating NILM model optimization, but they utilize compression approaches arbitrarily without considering the trade-off between model performance and computational cost. In this work, we present a NILM model optimization framework for edge deployment. The proposed edge optimization engine optimizes a NILM model for edge deployment depending on the edge device’s limitations and includes a novel performance-aware algorithm to reduce the model’s computational complexity. We validate our methodology on three edge application scenarios for four domestic appliances and four model architectures. Experimental results demonstrate that the proposed optimization approach can lead up to a 36.3% average reduction of model computational complexity and a 75% reduction of storage requirements.

Data Requirements for Applying Machine Learning to Energy Disaggregation

Energy disaggregation, or nonintrusive load monitoring (NILM), is a technology for separating a household’s aggregate electricity consumption information. Although this technology was developed in 1992, its practical usage and mass deployment have been rather limited, possibly because the commonly used datasets are not adequate for NILM research. In this study, we report the findings from a newly collected dataset that contains 10 Hz sampling data for 58 houses. The dataset not only contains the aggregate measurements, but also individual appliance measurements for three types of appliances. By applying three classification algorithms (vanilla DNN (Deep Neural Network), ML (Machine Learning) with feature engineering, and CNN (Convolutional Neural Network) with hyper-parameter tuning) and a recent regression algorithm (Subtask Gated Network) to the new dataset, we show that NILM performance can be significantly limited when the data sampling rate is too low or when the number of distinct houses in the dataset is too small. The well-known NILM datasets that are popular in the research community do not meet these requirements. Our results indicate that higher quality datasets should be used to expedite the progress of NILM research.

A generative model for non-Intrusive load monitoring in commercial buildings

In the recent years, there has been an increasing academic and industrial interest for analyzing the electrical consumption of commercial buildings. Whilst having similarities with the Non Intrusive Load Monitoring (NILM) tasks for residential buildings, the nature of the signals that are collected from large commercial buildings introduces additional difficulties to the NILM research causing existing NILM approaches to fail. On the other hand, the amount of publicly available datasets collected from commercial buildings is very limited, which makes the NILM research even more challenging for this type of large buildings. In this study, we aim at addressing these issues. We first present an extensive statistical analysis of both commercial and residential measurements from public and private datasets and show important differences. Secondly, we develop an algorithm for generating synthetic current data based on a modelization of the current flowing through an electrical device. We then demonstrate that our electrical device model fits well real measurements and that our simulations are realistic by using the quantitative metrics described in the previous section. Finally, to encourage research on commercial buildings we release a synthesized dataset called SHED that can be used to evaluate NILM algorithms.