NILMFormer: Non-Intrusive Load Monitoring that Accounts for Non-Stationarity

2025-06-06

AI Summary

To address distributional shifts within sliding windows caused by non-stationarity in real-world smart meter data for Non-Intrusive Load Monitoring (NILM), this paper proposes a Transformer-based non-stationarity modeling and correction framework. The method comprises two key innovations: (1) a subsequence stationarization/de-stationarization preprocessing mechanism that explicitly mitigates temporal distributional shift; and (2) a lightweight, timestamp-only time-aware positional encoding—replacing conventional periodic encodings—to enhance model robustness to non-stationary dynamics. Evaluated on four real-world datasets, the approach consistently outperforms existing state-of-the-art methods. It has been deployed as the core algorithm in Électricité de France’s (EDF) large-scale NILM service platform, serving millions of residential customers. The solution delivers high-accuracy, scalable household-level load disaggregation while maintaining computational efficiency and practical deployability.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Intelligent Robots: State Estimation
• Search and Optimization: Distributed Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Millions of smart meters have been deployed worldwide, collecting the total power consumed by individual households. Based on these data, electricity suppliers offer their clients energy monitoring solutions to provide feedback on the consumption of their individual appliances. Historically, such estimates have relied on statistical methods that use coarse-grained total monthly consumption and static customer data, such as appliance ownership. Non-Intrusive Load Monitoring (NILM) is the problem of disaggregating a household's collected total power consumption to retrieve the consumed power for individual appliances. Current state-of-the-art (SotA) solutions for NILM are based on deep-learning (DL) and operate on subsequences of an entire household consumption reading. However, the non-stationary nature of real-world smart meter data leads to a drift in the data distribution within each segmented window, which significantly affects model performance. This paper introduces NILMFormer, a Transformer-based architecture that incorporates a new subsequence stationarization/de-stationarization scheme to mitigate the distribution drift and that uses a novel positional encoding that relies only on the subsequence's timestamp information. Experiments with 4 real-world datasets show that NILMFormer significantly outperforms the SotA approaches. Our solution has been deployed as the backbone algorithm for EDF's (Electricit'e De France) consumption monitoring service, delivering detailed insights to millions of customers about their individual appliances' power consumption. This paper appeared in KDD 2025.

Problem

• Disaggregates household power usage into individual appliances.
• Addresses non-stationary data drift in smart meter readings.
• Improves accuracy of appliance-level energy consumption monitoring.

Innovation

• Transformer-based architecture for NILM
• Subsequence stationarization to reduce drift
• Timestamp-based positional encoding scheme