کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

Adaptive Loss-Drift Federated Learning for Non-Intrusive Load Monitoring

نویسندگان :

Ali Darvishi¹ Mohammad Reza Mansouri² Amir Reza Setayesh Matin³ Reza Gharibi⁴ Behnam Ranjbar⁵ Rahman Dashti⁶

1- Persian Gulf University 2- Persian Gulf University 3- Persian Gulf University 4- Persian Gulf University 5- University of Tehran 6- Persian Gulf University

کلمات کلیدی :

Non-intrusive Load monitoring (NILM)،Federated Learning،Optimization،Energy Disaggregation،Privacy،Artificial intelligence (AI)

چکیده :

Non-invasive load monitoring (NILM) is used to monitor household consumption at the appliance level, which is considered important for energy management and cost reduction. However, data confidentiality and privacy policies are viewed as major challenges in the training of NILM models. Federated learning (FL) is presented as a promising solution, as collaborative global training can be performed without raw data being exchanged. However, traditional FL aggregation methods such as Uniform averaging and its weighted forms are not capable of fully addressing user data heterogeneity, leading to weaker convergence and lower separation performance. In this paper, an adaptive method called Loss–Drift Federated Aggregation is presented, where dynamic weights are determined from three factors and assigned to each client: the amount of data, the local loss value, and the level of model drift. In this method, less weight is assigned to clients whose update behavior is observed to be unstable. Meanwhile, greater influence is given to clients whose behavior is more stable. A lightweight 1D-CNN model is used as the NILM predictor, making the approach suitable for deployment on edge devices. Experiments are performed on the REFIT dataset using a unified federated simulation setup consisting of 50 global rounds and 30 fine-tuning iterations. The results indicate that the proposed method yields consistent improvements over Uniform averaging and FedAvg-W, achieving up to 27% MAE reduction and showing greater robustness under heterogeneous conditions. These outcomes confirm the effectiveness of drift-aware adaptive aggregation for privacy preserving NILM.

لیست مقالات

لیست مقالات بایگانی شده

Integrating Distributed Renewables in Smart Urban Environments: A Smart Home Energy Management Approach

Masoud Alilou

Medial Frontal Network Dynamics Underlying Feedback Processing

Amin Mohammad Mohammadi - Zahra Ramezani - Farid Hosseinzadeh - Elias Ebrahimzadeh - Hamid Soltanian-Zadeh

Breast Cancer Detection by Time-Reversal Imaging Using Ultra-Wideband Modified Circular Patch Antenna Array

Mohammad Haghpanah - Zahra Ghattan Kashani - Atefeh Khalili Param

An Autonomous Multi Agent Q-Learning Approach for Resource Allocation in D2D-Enabled Heterogeneous Networks

Pouya Akhoundzadeh - Ghasem Mirjalily - Mohammad taghi Saadeghi

Brain Tumor Segmentation Using U-net and U-net++ Networks

Seyyed Ali Mortazavi-Zadeh - Alireza Amini - Hamid Soltanian-Zadeh

A Transformer-Based Model for Similar Fashion Image Retrieval with Image and Text Features

Zahra Sheykhvand - Milad Farzalizadeh - Majid Meghdadi

Designing a Suitable Antenna and Simple Receiver for Detection of Partial Discharge in the UHF Band

Seyed Hossein Kasaei - Mohammad Hamed Samimi

Design and fabrication tip tapered fiber optic dopamine sensor based on LSPR

Roksana Esmaeilpour - Mohammad Ismail zibaii - Masoumeh Barkand - Marzieh Pajouhandeh - Soroush Rostami - Mehdi Banihashemi - Mohammad-Mahdi Babakhani-fard

ارائه مبدل DC-DC غیر ایزوله هیبریدی بهره ولتاژ بالا با سوئیچ فعال سلفی

حسن زارعین - مجتبی حیدری - سیدمحمد دهقان دهنوی

Non-contact Radar Technology and Machine Learning for Automated Sleep Apnea-Hypopnea Syndrome Detection

ُSaman Faridsoltani - Mohaddeseh Sadeghi - Zahra Rahmani - Somayyeh Chamaani

بیشتر