Skip to main content
Disciplines
Automotive Business IT + Informatics Construction + Real Estate Electrical Engineering + Electronics Energy + Sustainability Insurance + Risk Finance + Banking Management + Leadership Marketing + Sales Mechanical Engineering + Materials
Events
DE
EN
Books
Journals
Topic Page
Marketing
Start single access now
Access for companies
EXTENDED SEARCH
Log in
Top

2020 | Book

Introduction Read chapter Read first chapter

Machine Learning Approaches to Non-Intrusive Load Monitoring

Authors: Dr. Roberto Bonfigli, Prof. Stefano Squartini

Publisher: Springer International Publishing

Book Series : SpringerBriefs in Energy

Part of: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Login to get access
insite
SEARCH

About this book

Research on Smart Grids has recently focused on the energy monitoring issue, with the objective of maximizing the user consumption awareness in building contexts on the one hand, and providing utilities with a detailed description of customer habits on the other. In particular, Non-Intrusive Load Monitoring (NILM), the subject of this book, represents one of the hottest topics in Smart Grid applications. NILM refers to those techniques aimed at decomposing the consumption-aggregated data acquired at a single point of measurement into the diverse consumption profiles of appliances operating in the electrical system under study.

This book provides a status report on the most promising NILM methods, with an overview of the publically available dataset on which the algorithm and experiments are based. Of the proposed methods, those based on the Hidden Markov Model (HMM) and the Deep Neural Network (DNN) are the best performing and most interesting from the future improvement point of view. One method from each category has been selected and the performance improvements achieved are described. Comparisons are made between the two reference techniques, and pros and cons are considered. In addition, performance improvements can be achieved when the reactive power component is exploited in addition to the active power consumption trace.

Table of Contents

Frontmatter
Chapter 1. Introduction
Abstract
In the recent years, the public awareness on energy saving themes has been constantly increasing. Indeed, the consequences of global warming are now tangible and studies have demonstrated that they are directly related to human activities and their inefficient use of energy and natural resources. The response of governments and public institutions to counteract this trend is to promote policies for reducing energy waste and intelligently use natural resources. The electricity grid is a key component in this scenario: the original electromechanical grid, where the information flow was one-directional, is transforming into the new digital smart grid where the information flows from the energy provider to distributed sensors and generator stations and vice versa. Part of this change involves the integration of smart meters in the grid in order to provide detailed consumption information both to the consumers and to the energy provider.
Roberto Bonfigli, Stefano Squartini
Chapter 2. Non-intrusive Load Monitoring
Abstract
The issues relating to the energy conservation and efficiency have gained a role of great importance, from the point of view of both the consumer and the energy provider. Furthermore, over the years, the infrastructures for energy distribution have undergone an ageing process, which have led to the study of the possibility in smart grids implementation, in which a set of information from detection and network management systems can be transmitted in addition to energy.
Useful information, about the characteristics and operating behaviour of an electrical system, can be obtained by means of the power consumption analysis, in order to predict the power demand (load forecasting), to apply management policies and to avoid overloading or blackouts over the energy network. Similarly, from the user perspective, the lifestyle of the people in a house can be predicted by the energy consumption analysis, allowing to implement policies for advantageous time tariffs.
Over the years, several studies have demonstrated that the energy consumption awareness (i.e., which appliances are operating at a certain time instant and how much electrical power they are consuming) influences the user behaviour. Specifically, the awareness conducts to moderate energy consumption, resulting in monetary savings and reduction of the energy required to the provider. Furthermore, applying this consideration to commercial or industrial environments, it may provide larger energy saving.
In the struggle to improve the energy efficiency of residential environments, the availability of information about the appliances in use can support automated optimization approaches.
Load monitoring has become a challenging problem, and several techniques have been studied to solve it. This work is focused on Non-Intrusive Load Monitoring (NILM) algorithms, which aim to separate the aggregated energy consumption signal, measured in a single centralized point, in the individual signals from each appliance, using a simple hardware but smart software algorithms. This solution replaces a distributed smart socket grid inside the house, resulting in lower implementation costs and less invasive solutions for the end user.
Roberto Bonfigli, Stefano Squartini
Chapter 3. Background
Abstract
The computers are able to perform complex calculus operations in a short amount of time. However computers cannot compete with humans in dealing with: common sense, ability to recognize people, objects, sounds, comprehension of natural language, ability to learn, categorize, generalize.
Therefore, why does the human brain show to be superior w.r.t common computers for these kind of problems? Is there any chance to mimic the mechanisms characterizing the way of working of our brain in order to produce more efficient machines?
In the field of signal analysis, the aim is the characterization of such real-world signals in terms of signal models, which can provide the basis for a theoretical description of a signal processing system. They are potentially capable of letting us learn a great deal about the signal source, without having to have the source available.
Therefore, in this chapter two families of modelling technique are described, i.e., the hidden Markov models (HMM) and the Deep Neural Network (DNN). After a theoretical description, the algorithms used for their parameter estimation are described, with a focus on the most widely model structure used in the field of the NILM.
Roberto Bonfigli, Stefano Squartini
Chapter 4. HMM Based Approach
Abstract
Approaches based on hidden Markov models (HMMs) have been devoted particular attention in the last years. AFAMAP (Additive Factorial Approximate Maximum a Posteriori) has been introduced in Kolter and Jaakkola to reduce the computational burden of FHMM. The algorithm bases its operation on additive and difference FHMM, and it constrains the posterior probability to require only one HMM change state at any given time.
Roberto Bonfigli, Stefano Squartini
Chapter 5. DNN Based Approach
Abstract
The recent success of Deep Neural Networks (DNN) in several application scenarios drove the scientific community to employ this paradigm also for NILM. Kelly and Knottenbelt compared three alternative DNNs: in the first, they employed a convolutional layer followed by long short-term memory (LSTM) layers to estimate the disaggregated signal from the aggregate one. In the second, a denoising autoencoder composed of convolutional and fully connected layers is trained to provide a denoised signal from the aggregate one. The third network estimates the start time, the end time and the mean power demand of each appliance. The algorithms were evaluated on the UK-DALE dataset and showed superior performance with respect to the combinatorial optimization and FHMM algorithms implemented in the Non-intrusive Load Monitoring Toolkit (NILMTK).
Roberto Bonfigli, Stefano Squartini
Chapter 6. Conclusions
Abstract
In this book, the Machine Learning approaches for Non-Intrusive Load Monitoring have been studied. Within all the techniques explored by the scientific community, this work has been focused on the hidden Markov model based and the deep neural network based, since their capability and promising performance at the forefront of the improvements could be introduced.
Roberto Bonfigli, Stefano Squartini
Backmatter
Metadata
Title
Machine Learning Approaches to Non-Intrusive Load Monitoring
Authors
Dr. Roberto Bonfigli
Prof. Stefano Squartini
Copyright Year
2020
Electronic ISBN
978-3-030-30782-0
Print ISBN
978-3-030-30781-3
DOI
https://doi.org/10.1007/978-3-030-30782-0