On-line Bayes Estimation of Renewable Energy Share in Modern Power Systems under an Autoregressive Model with Non-Gaussian Noise: A New Methodology

Estimating the renewable energy share is challenging because renewable generation is inherently stochastic. Numerous factors, including meteorological conditions, seasonal variations, electricity demand, market dynamics, and technological constraints, influence it. Statistical estimation methods that have emerged as tools for modelling the share of renewable energy from observational data can address these challenges. Bayesian statistical methods offer a complementary framework by incorporating prior knowledge and producing probabilistic estimates, unlike other estimation methods. In this paper, a new Bayesian estimation method for an autoregressive (AR) model was developed to estimate the Renewable Energy Source Share (RESS) when, as is common in data analysis of the Italian power system, the noise has a Logistic distribution. The