Search

Precise global solar radiation (GSR) measurements in a given location are very essential for designing and supervising solar energy systems. In the case of rarity or absence of these measurements, it is important to have a theoretical or empirical model to compute the GSR values. Therefore, the main goal of this work is to offer, to designers and engineers of solar energy systems, an appropriate and accurate way to predict the half-hour global solar radiation (HHGSR) time series from some available meteorological parameters (relative humidity, air temperature, wind speed, precipitation, and acquisition time vector in half-hour scale). For that purpose, two intelligent models are developed: the first one is a multivariate dynamic neural network with feedback connection, and the second is a multivariate static neural network. The database used to build these models was recorded in Agdal’s meteorological station in Marrakesh, Morocco, during the years of 2013 and 2014, and it was divided into two subsets. The first subset is used for training and validating the models, and the second subset is used for testing the efficiency and the robustness of the developed models. The obtained results, in terms of the statistical performance indicators, demonstrate the efficiency of the developed forecasting models to accurately predict the HHGSR parameter in the city of Marrakesh, Morocco. [ABSTRACT FROM AUTHOR]

The MIMO antenna is one of the essential elements for the implementation of the new fifth-generation (5 G) technologies. The design of the MIMO antenna goes through several steps; the important one is the design of the single-element antenna that resonates at the desired frequency band(s). Its design becomes more difficult if it is a multi-band antenna. In addition, determining the dimensions of multi-band antennas by simulation or mathematical models is both more complicated, difficult, and time-consuming. In this research, different approaches based on artificial neural networks (ANN), radial basis function neural networks (RBFNN), and adaptive neuro-fuzzy inference systems (ANFIS) are used to predict the dimensions of a dual-band circular patch antenna with defected ground structure (DGS) for mm-wave 5 G applications, and then the performances of the three models are compared. To determine the training and the testing data, 280 antennas with different patch values and added slots in the ground plane are simulated. The resonant frequencies of the dual-band circular patch antenna and the height of the substrate are used as the input vector for the ANN, RBFNN, and ANFIS models. Two hundred twenty simulated antennas were used for training, and the remaining 60 data points were used to test the ANN, RBFNN, and ANFIS models. The performance of these models is compared in the training and testing process using some statistical criteria such as MSE, MAE, and RMSE. The proposed dual-band circular patch single-element antenna has a compact size of 5 × 6 × 0.8 m m 3 and exhibits good performance, such as broad band (3.8 and 3 GHz), high gain (6 and 7.1 dB), and high efficiency (99 % and 95 % ) in both operating bands. After studying the single-element antenna and in order to improve its performance, we proposed a 4 × 4 MIMO antenna with a total volume of 17 × 17 × 0.8 m m 3 that was designed by using four copies of the single-element antenna arranged orthogonally on the same low-loss Rogers Duroid 5880 substrate. The suggested 4 × 4 MIMO antenna achieves a high level of isolation (more than - 30 dB in both operating bands), which is obtained by inserting a cross-shaped decoupling structure among elements on the upper and lower sides of the substrate. Moreover, the proposed MIMO antenna has a good gain of 6.4 and 7.5 dB in both bands and a high radiation efficiency of 99.5 % and 99 % at 28 and 38 GHz, respectively. Furthermore, the diversity performance of the MIMO antenna is evaluated using a variety of crucial indicators, all of which greatly surpass realistic norms. The envelope correlation coefficient (ECC) is less than 0.005, the diversity gain (DG) is greater than 9.99 dB, the total active reflection coefficient (TARC) is below 10 dB, and the channel capacity loss (CCL) is no longer than 0.4 bits/s/Hz. In addition, the proposed dual band circular patch single element antenna and the 4 × 4 MIMO antenna are fabricated, and the measured results are in good agreement with the simulated results using HFSS and CST software. [ABSTRACT FROM AUTHOR]

In this paper we propose to predict the notch frequency of an ultra-wideband (UWB) antenna which operates in the frequency band from 3.85 GHz to 12.38 GHz. The prediction of the notch frequency in order to avoid interferences between (WLAN) IEEE802.11a and HIPERLAN/2 WLAN applications and UWB technology is achieved using the artificial neural networks (ANN) technique. The developed ANN is optimized with the help of K-fold cross validation method which allows us to divide the datasets into 10 subsets in the training phase. The simulated datasets are generated by controlling high frequency structural simulator (HFSS) from MATLAB using a VB script. The performance of the ANN technique is assessed using some statistical criteria. During the training process, the mean absolute percentage error (MAPE) between the simulated and the predicted ANN notch frequencies is 0,125. A comparison between simulated, theoretical, and ANN results has been achieved during the test and validation process, good accuracy is obtained between the simulated and the ANN predictions. The proposed UWB antenna exhibits a notch band from 5.1 GHz to 6.0 GHz with a notch frequency of approximately 5.51 GHz. [ABSTRACT FROM AUTHOR]