# A Hybrid Federated Learning Framework for Enhancing Privacy and Robustness in Non-Intrusive Load Monitoring

## TL;DR

This paper introduces a hybrid federated learning framework to improve privacy and robustness in non-intrusive load monitoring for smart grids.

## Contribution

A novel hybrid framework dynamically switching between centralized and decentralized FL modes to enhance privacy and robustness in NILM.

## Key findings

- The proposed defense in centralized FL limits error metric increases to under 15.4% with 30% malicious clients.
- Decentralized FL with GAT-based topology reduces MAE by 17.2% after an attack and restores model performance quickly.
- The framework significantly improves robustness and privacy in smart-grid measurements using public datasets.

## Abstract

Non-intrusive load monitoring (NILM), as a key technology in smart-grid advanced metering infrastructure, aims to disaggregate mains power from smart meters into individual load-level power consumption. Traditional NILM methods require centralizing sensitive measurement data from users, which poses significant privacy risks. Federated learning (FL) enables collaborative training without centralized measurement data, effectively preserving privacy. However, FL-based NILM systems face serious threats from attacks such as model inversion and parameter poisoning, and rely heavily on the availability of a central server, whose failure may compromise measurement robustness. This paper proposes a hybrid FL framework that dynamically switches between centralized FL (CFL) and decentralized FL (DFL) modes, enhancing measurement privacy and system robustness simultaneously. In CFL mode, layer-sensitive pruning and robust parameter aggregation methods are developed to defend against model inversion and parameter poisoning attacks; even with 30% malicious clients, the proposed defense limits the increases in key error metrics to under 15.4%. In DFL mode, a graph attention network (GAT)-based dynamic topology adapts to mitigate topology poisoning attacks, achieving an approximately 17.2% reduction in MAE after an attack and rapidly restoring model performance. Extensive evaluations using public datasets demonstrate that the proposed framework significantly enhances the robustness of smart-grid measurements and effectively safeguards measurement privacy.