Towards Resilient Energy Forecasting: A Robust Optimization Approach

Abstract

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Resilient Electricity Load Forecasting Network with Collective Intelligence Predictor for Smart Cities

“…Robust Model [35] ✓ × Transformer [33,34] △ × Forecasting Network ✓ △ Federated Learning [38][39][40] × ✓ Boosting Ensemble Learning [36,37] ✓ × Temporal Convolutional Network [31,32] △ × Recurrent Neural Network Derivatives [30] △ ×…”

“…As well as relying on masking, padding, or imputation to make existing models capable of handling MV, several forecasting model designs could directly handle the MV. Stratigakos et al [35] proposed handling the MV with Linear Programming (LP) to formulate a robust regression model that minimizes the worst-case loss when a subset of the independent variables has MV. The authors noted that their method can handle up to 50% of MV.…”

Resilient Feature-driven Trading of Renewable Energy with Missing Data

“…Retraining typically outperforms impute-then-regress methods, however, the number of additional models required might render it impractical. To address this challenge, [8] develops a robust optimization approach to enable model resilience while also maintaining practicality for energy forecasting applications. Nonetheless, missing data at test time for renewable trading applications have not been thoroughly examined.…”

“…In this work, inspired by [6] and [8], we develop a featuredriven model to directly forecast the trading decisions of a renewable producer participating in a day-ahead electricity market that is resilient to missing data. We leverage robust optimization to formulate a model that optimizes for the worstcase trading cost when a subset of features is missing at test time.…”

Data is missing again -- Reconstruction of power generation data using k -Nearest Neighbors and spectral graph theory

“…patterns 7 , or need not even make any assumptions 8 , and estimate AR models. Other works develop models which are robust to missing data 9 . Another alternative is to provide imputations for missing values, i.e., to replace the data points which are missing with plausible values.…”