The tilt angle is one of the most important parameters for solar radiation collection in a photovoltaic panel, its optimization helps to increase the total solar radiation received. Numerous methods have been used to calculate the tilt angle around the world; unfortunately, many of these models are for fixed angles, which fail to adapt to the position of the sun throughout the year, resulting in the loss of significant amounts of irradiation that could be exploited. In this work we propose a series of optimal monthly tilt angles for the city of Coatzacoalcos, Veracruz, obtained by means of a neural network that considers day of the year, declination, and the clearness index as input variables. Its values range from 39.10 degrees in January, in June it has its lowest value with -4.88 and ends with values of 41.18 in December. Theoretical calculations of incident irradiation on a tilted plane are also made, using the isotropic Liu-Jordan model and the anisotropic Klucher model, to make comparisons of energy reception between a surface with a fixed tilt angle equal to the latitude (18.1378 degrees) and the proposed angles. The results show that using an isotropic method the irradiation obtained by panels with monthly angles increases 4.11% more and under the anisotropic method it increases 5.06% more. Thus, it is corroborated that it is possible to use neural networks to predict monthly tilt angles.
