Ricerc@Sapienza

Echo state networks (ESNs), belonging to the wider family of reservoir computing methods, are a powerful tool for the analysis of dynamic data. In an ESN, the input signal is fed to a fixed (possibly large) pool of interconnected neurons, whose state is then read by an adaptable layer to provide the output. This last layer is generally trained via a regularized linear least-squares procedure.

In this chapter we review two additional types of Recurrent Neural Network, which present important differences with respect to the architectures described so far. More specifically, we introduce the nonlinear auto-regressive with eXogenous inputs (NARX) neural network and the Echo State Network. Both these networks have been largely employed in Short Term Load Forecast applications and they have been shown to be more effective than other methods based on statistical models.

In this paper, we propose a novel data-driven approach for removing trends (detrending) from nonstationary, fractal and multifractal time series. We consider real-valued time series relative to measurements of an underlying dynamical system that evolves through time. We assume that such a dynamical process is predictable to a certain degree by means of a class of recurrent networks called Echo State Network (ESN), which are capable to model a generic dynamical process.

Forecasting the energy production of photovoltaic plants is today an essential tool for asset owners because it has direct economic implications on the net operating income of the plants whose generated energy is sold in competitive electricity markets. In this paper, we propose an innovative distributed decentralized prediction technique for the forecasting of power generated by several PV plants.

We propose a deep architecture for the classification of mul-tivariate time series. By means of a recurrent and untrained reservoir we generate a vectorial representation that embeds temporal relationships in the data. To improve the memorization capability, we implement a bidirectional reservoir, whose last state captures also past dependencies in the input. We apply dimensionality reduction to the final reservoir states to obtain compressed fixed size representations of the time series.

Renewable electricity generation has variable and non-dispatchable output that rises several technical, economic and feasibility concerns, calling for energy storage capacity and forecasting techniques to allow the integration of large amounts of variable generation into existing grids. These problems need careful attention in small islands that are not connected to the national transmission grid. In this paper, we present a study for the small Italian island of Ponza on the use of Echo State Networks to forecast real-world energy time series.