Non-Intrusive Load Monitoring Toolkit (nilmtk)

The super-state hidden Markov model disaggregator that uses a sparse Viterbi algorithm for decoding. This project contains the source code that was use for my IEEE Transactions on Smart Grid journal paper.

A repository of awesome Non-Intrusive Load Monitoring(NILM) with code.

An Non-Intrusive Load Disaggregation method based on Neural Network. A sequence-to-sequence model and a sequence-to-point model are proposed.

A reimplementation of Jack Kelly's rectangles neural network architecture based on Keras and the NILMToolkit.

⚠️ This repository is no longer actively maintained. It previously dealt with Non-Intrusive Load Monitoring (NILM), focusing on predicting household appliance status from aggregated power load data. We explored different thresholding methods and evaluated deep learning models for regression and classification tasks.

A new CNN architecture to perform detection, feature extraction, and multi-label classification of loads, in non-intrusive load monitoring (NILM) approaches, with a single model for high-frequency signals.

Overview of research papers with focus on low frequency NILM employing DNNs

Electrical Devices Identification Model (EDIM) for the identification of electrical devices by analyzing their energy consumption profiles.

An aided linear integer programming (ALIP) non-intrusive load monitoring (NILM) algorithm.

NILM performance evaluation functions use in the Springer Energy Efficiency journal paper.

ST-NILM is a new integrated architecture based on the Scattering Transform. It has a DCN (Deep Convolutional Network) with analytical wavelet-based non-trained weights, shared with fully connected output networks that perform event detection and multi-label classification of aggregate loads.

Supervised NILM using multiple-choice knapsack problem (MCKP).

Master's Dissertation: Unsupervised Low-Frequency NILM for Industrial Loads

The AMBAL-based NILM Trace generator (for NILMTK)

The C++ code for my Computing MSc energy disaggregation project

Complete set of electrical parameters, including energy consumption, power quality, and all harmonic contents (up to 2kHz). Time-synchronous data reported from real buildings and appliances at 1 second intervals.

WaveNILM as published at ICASSP 2019