TraceGAN: Synthesizing Appliance Power Signatures Using Generative Adversarial Networks

Abstract

Non-intrusive load monitoring (NILM) allows users and energy providers to gain insight into home appliance electricity consumption using only the building's smart meter. Most current techniques for NILM are trained using significant amounts of labeled appliances power data. The collection of such data is challenging, making data a major bottleneck in creating well generalizing NILM solutions. To help mitigate the data limitations, we present the first truly synthetic appliance power signature generator. Our solution, PowerGAN, is based on conditional, progressively growing, 1-D Wasserstein generative adversarial network (GAN). Using PowerGAN, we are able to synthesise truly random and realistic appliance power data signatures. We evaluate the samples generated by PowerGAN in a qualitative way as well as numerically by using traditional GAN evaluation methods such as the Inception score.

Construction of Sports Training Performance Prediction Model Based on a Generative Adversarial Deep Neural Network Algorithm

“…Sports data are rarely applied to good data mining and feature selection techniques and even less applied to study the impact of sports on physical index data [ 5 ]. In related research at home and abroad, the ID3 algorithm, an algorithm of decision tree analysis, is applied to study human grip muscle strength test data, and root nodes of different test index parameters are determined to get the indexes that can scientifically evaluate human muscle strength.…”

Orchestrating Heterogeneous Devices and AI Services as Virtual Sensors for Secure Cloud-Based IoT Applications

“…For single patterns and programmatic patterns, we use a generative adversarial network (GAN), to be explicit, a conditional Wasserstein GAN with gradient penalty (cWGAN-GP), comparable to [14]. The WGAN-GP framework gives us the possibility of generating realistic data based on the distribution of the training data [15][16][17].…”

“…As for the generation of synthetic data, TraceGAN [14] is a solution to generate appliance power using a GAN. It can be used to generate slightly different energy patterns for training NILM applications.…”

Recent trends of smart nonintrusive load monitoring in buildings: A review, open challenges, and future directions

“…1 For example, power consumption in residential households could be facilely decreased by more than 10% via providing end-users with their energy usage feedback; precisely, the way they consume their energy and their domestic devices' energy consumption throughout the day. 2 Consequently, developing robust solutions to offer end-users this information and motivate them to endorse more sustainable energy behaviors is becoming a hot research topic. 3 To that end, a straightforward strategy to achieve that objective is through adopting "intrusive load monitoring (ILM)," which involves the installation of individual smart meters/sensors at every single appliance in a building, which unfortunately comes with a high cost.…”

“…Collecting appliance‐level power consumption data is tremendously valuable to promote energy‐saving behavior and intelligently manage the demand response exchange between the utility and end‐users 1 . For example, power consumption in residential households could be facilely decreased by more than 10% via providing end‐users with their energy usage feedback; precisely, the way they consume their energy and their domestic devices' energy consumption throughout the day 2 . Consequently, developing robust solutions to offer end‐users this information and motivate them to endorse more sustainable energy behaviors is becoming a hot research topic 3 .…”