Privacy-Preserving Probabilistic Voltage Forecasting in Local Energy Communities

This paper presents a new privacy-preserving framework for the short-term (multi-horizon) probabilistic forecasting of nodal voltages in local energy communities. task is indeed becoming increasingly important cost-effectively managing network constraints context massive integration distributed resources. However, traditional tasks are carried out centrally, by gathering raw data end-users single database that exposes their private information. To avoid such privacy issues, this work relies on learning scheme, known as federated wherein individuals’ kept decentralized. The procedure then augmented with differential privacy, which offers formal guarantees trained model cannot be reversed-engineered to infer sensitive Moreover, problem framed using cross-series learning, allows smoothly integrate any client joining community (i.e., cold-start forecasting) without being plagued scarcity. Outcomes show proposed approach achieves improved performance compared non-collaborative (locally trained) models, and able reach trade-off between different architectures deep networks.