Application of AI for Short-Term PV Generation Forecast.

The efficient use of the photovoltaic power requires a good estimation of the PV generation. That is why the use of good techniques for forecast is necessary. In this research paper, Long Short-Term Memory, Bidirectional Long Short-Term Memory and the Temporal convolutional network are studied in depth to forecast the photovoltaic power, voltage and efficiency of a 1320 Wp amorphous plant installed in the Technology Support Centre in the University Rey Juan Carlos, Madrid (Spain). The accuracy of these techniques are compared using experimental data along one year, applying 1 timestep or 15 min and 96 step times or 24 h, showing that TCN exhibits outstanding performance, compared with the two other techniques. For instance, it presents better results in all forecast variables and both forecast horizons, achieving an overall Mean Squared Error (MSE) of 0.0024 for 15 min forecasts and 0.0058 for 24 h forecasts. In addition, the sensitivity analyses for the TCN technique is performed and shows that the accuracy is reduced as the forecast horizon increases and that the 6 months of dataset is sufficient to obtain an adequate result with an MSE value of 0.0080 and a coefficient of determination of 0.90 in the worst scenarios (24 h of forecast).

Design and analysis of recurrent neural networks for ultrafast optical pulse nonlinear propagation.

In this work, we analyze different types of recurrent neural networks (RNNs) working under several different parameters to best model the nonlinear optical dynamics of pulse propagation. Here we studied the propagation of picosecond and femtosecond pulses under distinct initial conditions going through 13 m of a highly nonlinear fiber and demonstrated the application of two RNNs returning error metrics such as normalized root mean squared error (NRMSE) as low as 9%. Those results were further extended for a dataset outside the initial pulse conditions used on the RNN training, and the best-proposed network was still able to achieve a NRMSE below 14%. We believe that this study can contribute to a better understanding of building RNNs employed for modeling nonlinear optical pulse propagation and of how the peak power and nonlinearity affect the prediction error.

An Optimized Self-Compensated Solution for Temperature and Strain Cross-Sensitivity in FBG Interrogators Based on Edge Filter.

Optical fiber sensors based on fiber Bragg gratings (FBGs) are prone to measurement errors if the cross-sensitivity between temperature and strain is not properly considered. This paper describes a self-compensated technique for canceling the undesired influence of temperature in strain measurement. An edge-filter-based interrogator is proposed and the central peaks of two FBGs (sensor and reference) are matched with the positive and negative slopes of a Fabry-Perot interferometer that acts as an optical filter. A tuning process performed by the grey wolf optimizer (GWO) algorithm is required to determine the optimal spectral characteristics of each FBG. The interrogation range is not compromised by the proposed technique, being determined by the spectral characteristics of the optical filter in accordance with the traditional edge-filtering interrogation. Simulations show that, by employing FBGs with optimal characteristics, temperature variations of 30 °C led to an average relative error of 3.4% for strain measurements up to 700µÏµ. The proposed technique was experimentally tested under non-ideal conditions: two FBGs with spectral characteristics different from the optimized results were used. The temperature sensibility decreased by 50.8% as compared to a temperature uncompensated interrogation system based on an edge filter. The non-ideal experimental conditions were simulated and the maximum error between theoretical and experimental data was 5.79%, proving that the results from simulation and experimentation are compatible.