Abstract

Abstract
Cooperation between different data owners may lead to an improvement in forecast quality—for instance, by benefiting from spatiotemporal dependencies in geographically distributed time series. Due to business competitive factors and personal data protection concerns, however, said data owners might be unwilling to share their data. Interest in collaborative privacy-preserving forecasting is thus increasing. This paper analyzes the state-of-the-art and unveils several shortcomings of existing methods in guaranteeing data privacy when employing vector autoregressive models. The methods are divided into three groups: data transformation, secure multi-party computations, and decomposition methods. The analysis shows that state-of-the-art techniques have limitations in preserving data privacy, such as (i) the necessary trade-off between privacy and forecasting accuracy, empirically evaluated through simulations and real-world experiments based on solar data; and (ii) iterative model fitting processes, which reveal data after a number of iterations. © 2020 International Institute of Forecasters

Abstract
Probabilistic forecasting of distribution tails (i.e., quantiles below 0.05 and above 0.95) is challenging for non-parametric approaches since data for extreme events are scarce. A poor forecast of extreme quantiles can have a high impact in various power system decision-aid problems. An alternative approach more robust to data sparsity is extreme value theory (EVT), which uses parametric functions for modelling distribution's tails. In this work, we apply conditional EVT estimators to historical data by directly combining gradient boosting trees with a truncated generalized Pareto distribution. The parametric function parameters are conditioned by covariates such as wind speed or direction from a numerical weather predictions grid. The results for a wind power plant located in Galicia, Spain, show that the proposed method outperforms state-of-the-art methods in terms of quantile score. © 2020 Elsevier B.V.

Abstract

Abstract
This paper presents a generic framework to evaluate and compare the quality of the uncertainties provided by probabilistic forecasts of power system state when used to perform security assessment for branch overloads. Besides exploiting advanced univariate and multivariate metrics that are traditionally used in weather prediction, the evaluation is complemented by assessing the benefits from exploiting probabilistic forecasts over the current practices of using deterministic forecasts of the system operating conditions. Another important feature of this framework is the provision of parameters tuning when applying flow probabilistic forecasts to perform security assessment for branch overloads. The quality and scalability of this framework is demonstrated and validated on recent historical data of the French transmission system. Although being developed to address branch overload problems, with proper adaptations, this work can be extended to other power system security problems. © 2019 Elsevier Ltd