Non-Intrusive Load Monitoring Model Based on SimCLR and Visualized Color V-I Trajectories

TL;DR

A new model for non-intrusive load monitoring uses self-supervised learning and domain adaptation to accurately identify appliances with minimal labeled data.

Contribution

A novel self-supervised framework combining SimCLR and adversarial domain adaptation for cross-domain appliance recognition.

Findings

The model achieved an F1-score of 0.9498 using only 10% labeled target data.

It outperformed supervised models trained on 30% data in cross-domain identification tasks.

Abstract

What are the main findings?
A novel self-supervised framework integrating SimCLR with adversarial domain adaptation effectively aligns cross-domain feature distributions using visualized color V-I trajectories.The proposed model achieved an F1-score of 0.9498 with only 10% labeled target data, surpassing the performance of supervised models trained on 30% data.

A novel self-supervised framework integrating SimCLR with adversarial domain adaptation effectively aligns cross-domain feature distributions using visualized color V-I trajectories.

The proposed model achieved an F1-score of 0.9498 with only 10% labeled target data, surpassing the performance of supervised models trained on 30% data.

What are the implications of the main findings?
Integrating adversarial mechanisms into self-supervised learning significantly mitigates domain shift challenges, ensuring robust appliance recognition across diverse household environments.The method drastically reduces dependence on manual data annotation, offering a cost-effective and scalable solution for deploying non-intrusive load monitoring systems in smart grids.

Integrating adversarial mechanisms into self-supervised learning significantly mitigates domain shift challenges, ensuring robust appliance recognition across diverse household environments.

The method drastically reduces dependence on manual data annotation, offering a cost-effective and scalable solution for deploying non-intrusive load monitoring systems in smart grids.

Current non-intrusive load monitoring (NILM) methods rely on large amounts of labeled historical data and face domain shift issues, which limits the application of deep learning models in practical scenarios. To this end, this paper proposes a SimCLR-ADA-LM framework based on visualized color V-I trajectories. Initially, unlabeled load data from the source domain (PLAID) and target domain (WHITED) are converted into RGB color V-I trajectories and input into the model. The framework enhances intra-class aggregation through contrastive learning and achieves inter-domain feature alignment via adversarial training between the encoder and the domain discriminator to obtain domain-invariant features. Subsequently, the model is fine-tuned using a small amount of labeled data from the target domain to achieve load identification. Ablation and comparative experimental results demonstrate that the proposed model exhibits superior performance advantages over conventional models in cross-domain identification tasks. Furthermore, it maintains significant learning efficiency and recognition robustness even under conditions of limited labeled data.

Figures