[MAR-EPF] Mixed Causal Non-Causal model for Electricity Price Forecasting
Ente: EC
Scadenza: 2029-08-31
Importo max: 247.553 EUR
Paese: EU
Descrizione
The project focuses on the development of a novel econometric methodology for electricity price forecasting (EPF), with the goal of improving existing methods. Forecasting electricity prices is crucial for energy market trading and stablity. However, the modeling and forecasting of these prices are complex due to the non-storability, the dependence on the transmission grid, and weather patterns. These features, joint with the green transition and expanding power supply from renewable energy sources, give rise to extreme events, heavy tails and complex dynamics in general. In this sense, better forecasting of electricity prices would lead to an improved risk assessment and monitoring for economic agents and engineering involved in the logistics of the electricity market.
To achieve this goal, the project relies on two main contributions: the use of mixed causal–non-causal (MAR) models and the explicit incorporation of physical constraints, such as distribution grid limitations and weather-driven effects, into the econometric framework. MAR models are a class of heavy-tailed autoregressive models that allows a process to depend also on its future. The forward looking specification of MAR models allows the framework to model non-linear and explosive dynamics while maintaining the linear and parsimonious specification typical of autoregressive models. This approach is novel in the literature for electricity price modeling, offering a bridge between the non-linear methods as Artificial Neural Networks (ANN) and traditional linear autoregressive models (ARMA) that are currently in use.
Moreover, by explicitly modeling physical constraints, including transmission network bottlenecks, generation capacity limits, and the influence of weather variables such as temperature, wind, and solar irradiance, the proposed framework will capture the underlying mechanisms that drive price formation in electricity markets, enhancing interpretability and improving forecasting accuracy.
Settori: Horizon Europe Topics
Vai al bando originale
Registrati gratis su Bandolo per trovare bandi compatibili con la tua azienda.