BINARY SHADING CLASSIFICATION BY MEANS OF AN IDENTIFICATION ALGORITHM FOR RADIATION TRANSITIONS USING LOW COST SENSORS

Resumo

In the current energy scenario, renewable energies such as solar and wind power have
significant relevance. In this context, an algorithm capable of recognizing irradiation transitions was
implemented. This algorithm was used to classify data obtained from light dependent resistance in two
classes: shadow and non-shadow. These results were compared with those obtained by a standard
algorithm that has a input global or beam irradiation values collected by pyranometer and pyreliometer
sensors, respectively. This way, a practical tool was obtained. Such tool uses data collected by low cost
sensors to perform an initial statistical study about days and hours when there is a higher level of
irradiation. Therefore, this information can be used to optimize the solar power capture. The irradiation
transition identification algorithm, generally, had a better compilance at peaks hours, where for the same
days it reached ranging around 100% and 87.10% using the global and beam components as input to the
standard algorithm, respectively. From the results, it is clears that a better performance of the irradiation
transition identification algorithm when using global irradiation values as input to the standard
algorithm, as well as a lower compilance of the irradiation transition algorithm during the dawn and
twilight periods.