This paper studies the role of model initialization in federated STLF, and proposes two initialization strategies from global and local perspectives, which effectively improve forecasting performance, as evidenced by reduced client drift, improved convergence behavior, and lower forecasting errors.
Abstract
Short-term load forecasting (STLF) provides essential information for numerous applications in modern power systems. However, accurate STLF often relies on fine-grained smart-meter data from distributed users, raising increasing concerns about data privacy. Federated learning (FL) has therefore emerged as a promising privacy-preserving paradigm for STLF. Nevertheless, this paper reveals structured heterogeneity in clients'load data. Specifically, clients exhibit different responses to exogenous factors and distinct temporal load profiles, which can degrade forecasting performance in FL. To mitigate these issues, this paper studies the role of model initialization in federated STLF, and proposes two initialization strategies from global and local perspectives. For global model initialization, when auxiliary public load data are available, a pretrained initialization strategy is developed to initialize the global model before federated training, thereby reducing client drift during the training process. For local model initialization, we propose SLIAvg, a sequential local initialization strategy that promotes a more consistent training process by allowing participating clients to start from progressively adapted models within each communication round. Since the proposed strategies only modify the initialization process, they are compatible with most existing FL frameworks and privacy-enhancing techniques. Experiments on real smart-meter data with two representative forecasting architectures demonstrate that the proposed strategies effectively improve forecasting performance, as evidenced by reduced client drift, improved convergence behavior, and lower forecasting errors.
Experimental results on real-world energy consumption datasets demonstrate that the proposed FL framework achieves competitive forecasting accuracy while preserving client data privacy, and a rigorous comparative analysis reveals that FedProx and FedTrimmedAvg consistently outperform FedAvg under non-IID conditions.
A. Tibermacine, Ilyes Naidji, Imad Eddine Tibermacine et al.· Frontiers in Energy Research· 1 citation
Modern power systems are growing increasingly complex as they integrate diverse generation sources to meet rising demand, making accurate load forecasting challenging. Recent advances in time-series foundation models (TSFMs) resulted in promising performance in zero-shot univariate load forecasting tasks. However, real-world load forecasting often involves multiple target variables and requires the integration of exogenous variables, raising important questions about the utility of TSFMs in realistic settings. In this study, we position Chronos-2, a recently developed model by Amazon, as a representative multi-channel TSFM that supports univariate, multivariate, and covariate-informed forecasting, and conduct a systematic investigation of how such models can be used for real-world load forecasting. While prior work has evaluated Chronos-2 on a limited number of energy-related tasks in a zero-shot setting, its performance relative to established task-specific deep learning models and its behavior when adapted using task-specific historical data remains insufficiently understood. In this work, we evaluate Chronos-2 on two real-world utility datasets, ISO New England and ENTSO-E, and benchmark it against widely used task-specific deep learning models. Our results show that Chronos-2 benefits substantially from task-specific fine-tuning and achieves strong short-horizon forecasting performance, but its zero-shot accuracy lags behind task-specific models and its forecasting error grows more rapidly with increasing forecast steps. Overall, this study provides a detailed characterization of the strengths and limitations of TSFMs such as Chronos-2 in grid load forecasting and offers practical insights into how a pretrained TSFM can be effectively adapted for operational load forecasting applications.
Varsha Pendyala, Yiwei Fu, Weizhong Yan et al.· 0 citations
The increasing integration of renewable energy is a key pathway toward sustainable energy transitions, but it also introduces supply-side uncertainty that makes demand flexibility management increasingly important. In this context, accurate load forecasting becomes a key task for balancing uncertain supply with demand-side behavior and supporting more reliable energy management. Most existing studies approach load forecasting by developing advanced forecasting models. However, in real-world commercial-user forecasting, load data often exhibit inter-user heterogeneity and temporal dynamics, making forecasting performance depend not only on model architecture, but also on how models are trained, shared, combined, and used; we refer to these design choices as forecasting strategies. Therefore, we study day-ahead individual load forecasting from a strategy-level perspective using historical load observations only. We compare homogeneous strategies, inter-user-heterogeneity-aware strategies, and temporal-dynamics-aware strategies using three real-world electricity consumption datasets from China. The results show that forecasting strategy design substantially affects prediction performance, and that both inter-user heterogeneity and temporal dynamics provide useful and complementary information for improving day-ahead individual load forecasting. We find that appropriate strategy design can improve MSE by more than 6% compared with using only a single model. Overall, the study provides useful methodological tools and practical insights for load forecasting and sustainable energy management.
Haoheng Qin, Liang Xu, Garrett Ge Gao et al.· Sustainability· 0 citations
Federated adaptation of time-series foundation models (TSFMs) is attractive for building energy forecasting because meter data are private, distributed, and highly non-IID. However, a single parameter-sharing strategy is unlikely to serve all pretrained TSFMs or building clients: fully shared adapters can suppress building-specific temporal behavior, while fully local adaptation discards cross-building transfer. We propose a personalized federated sparse adaptation framework with a heterogeneous temporal mixture-of-experts (MoE) adapter placed after the pretrained TSFM representation. A sequence-level router maps each 168-hour context window to a top-$k$ subset of experts specialized for periodicity, long-range interactions, local variation, trend-residual structure, and multi-resolution behavior. We compare global FL, local training, and personalized FL variants with globally shared or client-private expert banks. Across 50 buildings and three TSFM backbones, personalization consistently outperforms Global FL-MoE and Local MoE, while the best sparse-adaptation strategy varies by backbone and metric. Routing behavior further reveals client-level expert specialization, expert concentration, and near-uniform routing across backbones, showing that federated TSFM adaptation should be both client-aware and backbone-aware.
Priyanka Nihalchandani, Naman Srivastava, Varun Ojha et al.· 0 citations
A weeklong summer workshop brought higher education faculty to campus to explore how AI and machine learning materials can be adapted for their classrooms.
