Prediction system based on domotic weather sensors for the energy production of solar power plants
DOI:
https://doi.org/10.24084/repqj09.633Abstract
The prediction of the electrical energy generated by a photo-voltaic system is useful for estimating the profitability analysis of a project, without the need of expensive photovoltaic proto-types. Prediction systems are usually based on simulating the physical process of a photovoltaic module, under standard or average local weather conditions. These predictions introduce some errors caused by the use of a theoretical model or average climate data.In our investigations, we noted that the energy generated by a photovoltaic system is proportional to the cumulative measure-ment of the sun illuminance that is provided by a low-cost domo-tic weather station. From this experimental observation, this paper proposes a hardware/software system for predicting the electrical energy generated by a photovoltaic system, such as those existing in buildings. The hardware consists of a domotic installation for monitoring both electric energy and climate parameters. The software consists of a calibration procedure, which provides a proportional factor between sun illuminance and the energy production per unit of surface area of the photo-voltaic modules. Once the calibration procedure is completed, the photovoltaic energy production is predicted by factoring the sun illuminance provided by the weather station and the propor-tional factor provided by the calibration process. This method has been tested under real conditions and the accuracy reached up to 99.7% with an average value of 96.3%.