Víctor Martínez-Moll

FPGA-based stochastic echo state networks for time-series forecasting

Hardware implementation of artificial neural networks (ANNs) allows exploiting the inherent parallelism of these systems. Nevertheless, they require a large amount of resources in terms of area and power dissipation. Recently, Reservoir Computing (RC) has arisen as a strategic technique to design recurrent neural networks (RNNs) with simple learning capabilities. In this work, we show a new approach to implement RC systems with digital gates. The proposed method is based on the use of probabilistic computing concepts to reduce the hardware required to implement different arithmetic operations. The result is the development of a highly functional system with low hardware resources. The presented methodology is applied to chaotic time-series forecasting.

Optical characterization of a fixed mirror solar concentrator prototype by the ray-tracing procedure

The Fixed Mirror Solar Concentrator (FMSC) is a solar concentrator with static reflector and moving receiver whose design emerged in the seventies as an effort to reduce electricity production costs in solar thermal power plants. Solar concentrators based on this geometry were constructed in the seventies and eighties. A review of these prototypes is presented, highlighting the lack in research probably brought about by halting these projects, and the two main flaws of this research: poor theoretical analysis of the FMSC geometry and the unfortunate choice of the zero-width mirror limit hypothesis. In this paper, another methodology is presented to evaluate the FMSC geometry behavior: the construction of a FMSC prototype using lightweight materials with finite-width mirrors (nine mirrors) and an optical characterization by 3D ray-tracing tools. The results show good concordance between simulated and experimental data, showing that FMSC prototypes can be characterized optically by accurate ray-tracing tools.

Short-Term Spanish Aggregated Solar Energy Forecast

This work presents and compare six short-term forecasting methods for hourly aggregated solar generation. The methods forecast one day ahead hourly values of Spanish solar generation. Three of the models are based on MLP network and the other three are based on NARX. The two different types of NN use to forecast the same NWP data, comprising solar radiation, solar irradiation and the cloudiness index weighted with the installed solar power for the whole country. In addition of the NWP data the models are fed with the aggregated solar energy generation in hourly step given by the System Operator.

Low-cost hardware implementation of Reservoir Computers

The hardware implementation of massive Recurrent Neural Networks to efficiently perform time dependent signal processing is an active field of research. In this work we review the basic principles of stochastic logic and its application to the hardware implementation of Neural Networks. In particular, we focus on the implementation of the recently introduced Reservoir Computer architecture. We show the functionality and low hardware resources used to implement the Reservoir Computer by synthesizing a network performing a mathematical regression.

Stochastic-Based Implementation of Reservoir Computers

Hardware implementations of Artificial Neural Networks (ANNs) allow to exploit the inherent parallelism of these architectures. Nevertheless, ANN hardware implementation requires a large amount of hardware resources. Recently, Reservoir computing (RC) has arisen as an advantageous technique to implement Recurrent Neural Networks RNNs). In this work, we present an efficient approach to implement RC systems. The proposed methodology employs probabilistic logic to reduce the hardware area required to implement the arithmetic operations present in neural networks and conventional binary logic for the nonlinear activation function. We show the functionality and low hardware resources used by the proposed methodology.