Deep Domain Adaptation for Non-Intrusive Load Monitoring Based on a Knowledge Transfer Learning Network

Lin, Jun; Jin, Ma; Zhu, Jun; Liang, Huishi

doi:10.1109/tsg.2021.3115910

Cited by 50 publications

(14 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another key issue impeding the development of effective and reliable DTL-based ASR strategies is the large diversity of mathematical formulations utilized to describe the background of DTL-based SRT techniques. For example, in [162] Hu et al promote the idea of heterogeneous DTL while Fan et al [169] opt for statistical investigations of DTL-based methodologies, and [179][180][181] concentrate on deep DA. Despite that the studies reviewed in this article share the same DTL idea, they differ in their definitions and implementations based on the considered scenarios.…”

Section: Unification Of Dtlmentioning

confidence: 99%

Video surveillance using deep transfer learning and deep domain adaptation: Towards better generalization

Himeur

Al‐Maadeed

Kheddar

et al. 2023

Engineering Applications of Artificial Intelligence

View full text Add to dashboard Cite

Section: Unification Of Dtlmentioning

confidence: 99%

Video surveillance using deep transfer learning and deep domain adaptation: Towards better generalization

Himeur

Al‐Maadeed

Kheddar

et al. 2023

Engineering Applications of Artificial Intelligence

View full text Add to dashboard Cite

“…Second, a probabilistic neural network (PNN) is utilized to classify devices and transfer the knowledge between appliances. Lin et al 127 propose a TL‐based NILM using a TCNN model for learning the dynamic characteristics of individual devices' loads. Specifically, a domain adaption loss has been adopted for quantifying the domain distribution discrepancy between source and target domain representations.…”

Section: Overview Of Recent Smart Nilm Algorithmsmentioning

confidence: 99%

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

Himeur

Alsalemi

Bensaali

et al. 2022

Int J of Intelligent Sys

View full text Add to dashboard Cite

Smart nonintrusive load monitoring (NILM) represents a cost-efficient technology for observing power usage in buildings. It tackles several challenges in transitioning into a more effective, sustainable, and digital energy efficiency environment. This paper presents a comprehensive review of recent trends in the NILM field, in which we propose a multiperspective classification of existing smart NILM techniques. More attention is devoted to describing the contributions of deep learning, feature extraction, computing platforms, and application scenarios for NILM development. Accordingly, NILM technical aspects are first Int J Intell Syst. 2022;37:7124-7179. wileyonlinelibrary.com/journal/int investigated, including data collection devices and public data sets. Next, event-based and non-eventbased NILM algorithms are overviewed. Furthermore, potential limitations of existing solutions are identified, highlighting their technical challenges, especially those related to security and privacy preservation, data scarcity, results reproduction, and implementation and business difficulties. Lastly, future directions are explored to overcome the identified limitations.

show abstract

“…In order to overcome the labeled data bottleneck in NILM, researchers and developers are developing various techniques such as transfer learning [15], [31], [32], domain adaptation [33], [34], synthetic appliance power data generation [35], [36], Semi-supervised learning (SSL) techniques [37]- [39]. The world of SSL techniques is a particularly fascinating field, which could be the bridge between label-heavy supervised learning and a future of data-efficient modeling.…”

Section: Introductionmentioning

confidence: 99%

Non-intrusive load identification based on semi-supervised learning with flexible threshold and V-I trajectory signature

Tang

et al. 2023

Preprint

View full text Add to dashboard Cite

Non-intrusive load monitoring (NILM) can obtain fine-grained electricity consumption information for each appliance in the user's home without installing additional hardware sensors. As deep learning has grown quickly, many deep learning algorithms have been used to solve NILM problems, and have achieved good load identification results. However, many supervised learning algorithms need to be based on a large amount of labeled data, which is time-consuming and laborious to obtain, and difficult to implement in actual deployment. In this paper, we propose a semi-supervised learning algorithm that combines consistency regularization and pseudo-labels to help identify appliances with a small amount of labeled data and some unlabeled data. In addition, given the different learning difficulties of different appliances among users, for example, the feature learning difficulty of multi-state appliances is often higher than that of single-state appliances. Therefore, we employ that the threshold values of different appliance classes be changed in a flexible way at each time step so that the informative unlabeled data and their pseudo-labels can be passed. We have carried out simulation experiments on public dataset, and the experimental results show that the proposed algorithm achieves better appliance identification results than the state-of-the-art methods.

show abstract

Deep Domain Adaptation for Non-Intrusive Load Monitoring Based on a Knowledge Transfer Learning Network

Cited by 50 publications

References 29 publications

Video surveillance using deep transfer learning and deep domain adaptation: Towards better generalization

Video surveillance using deep transfer learning and deep domain adaptation: Towards better generalization

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

Non-intrusive load identification based on semi-supervised learning with flexible threshold and V-I trajectory signature

Contact Info

Product

Resources

About