Fed‐NILM: A federated learning‐based non‐intrusive load monitoring method for privacy‐protection

Abstract

Non-intrusive load monitoring (NILM) is essential for understanding consumer power consumption patterns and may have wide applications such as in carbon emission reduction and energy conservation. Determining NILM models requires massive load data containing different types of appliances. However, inadequate load data and the risk of power consumer privacy breaches may be encountered by local data owners when determining the NILM model. To address these problems, a novel NILM method based on federated learning (FL) called Fed-NILM is proposed. In Fed-NILM, instead of local load data, local model parameters are shared among multiple data owners. The global NILM model is obtained by averaging the parameters with the appropriate weights. Experiments based on two measured load datasets are performed to explore the generalization capability of Fed-NILM. In addition, a comparison of Fed-NILM with locally trained NILM models and the centrally trained NILM model is conducted. Experimental results show that the Fed-NILM exhibits superior performance in terms of scalability and convergence. Fed-NILM out performs locally trained NILM models operated by local data owners and approaches the centrally trained NILM model, which is trained on the entire load dataset without privacy protection. The proposed Fed-NILM significantly improves the co-modelling capabilities of local data owners while protecting the privacy of power consumers.

DP$^2$-NILM: A Distributed and Privacy-preserving Framework for Non-intrusive Load Monitoring

“…On the other hand, the emergence of federated learning [26] not only provides privacy guarantees for smart meter data but solves the challenge of data isolation, which therefore brings considerable benefits to DNN-based NILM models. Despite its promising future, applying FL to NILM has only received attention very recently [27,28,29,30]. For instance, [27] proposed a FederatedNILM framework to enable NILM task in the FL paradigm at the residential level.…”

“…For instance, [27] proposed a FederatedNILM framework to enable NILM task in the FL paradigm at the residential level. [28] utilized the FL paradigm to improve the model performance for NILM in both residential and industrial scenarios. [29] proposed a FedNILM framework utilizing model compression to reduce the computation overhead while retaining satisfying performance for NILM.…”

Federated office plug-load identification for building management systems

“…The evaluation of hardware metrics during FL training is not well investigated in research. Other papers focus on model performance on simulated nodes [9,16,35,68,71,76,77,80]. These setups do not cover the system behaviour under real-world conditions.…”

“…The focus is on household applications and stationary equipment [68]. Exceptions are Yang et .al [80] who ran experiments on mobile devices and Wan et al [77] who used a data set of heavy-machinery from a Brazilian poultry feed factory. In both cases, they clustered the data to have similar properties.…”

On enabling collaborative non-intrusive load monitoring for sustainable smart cities

“…In recent years, federated learning was proposed 13 to train a global model collaboratively without exchanging the raw data of stakeholders. The existing NILM federated learning solutions are deep learning oriented in a centralised setting 14 – 16 . The central server coordinates all the stakeholders to train a neural network model.…”