Image Processing Based Dust Detection and prediction of Power using ANN in PV systems

Currently in the market, the most effective solar panels constitute the efficiency ratings as high as 22.8%, while majority of the panel efficiencies vary from 15% to 17%. However, the theoretical photovoltaic conversion efficiency reaches 86.6% [1]. This is mainly due to the fact that, it is assumed that each photon is optimally used and have high concentration ratio which is not the case for terrestrial solar panels and hence have higher efficiencies. On the other hand, different factors disturb the power output of a terrestrial module in which most of the factors are related to the module itself. In the UAE region, one of the key issues that affect the performance of the photovoltaic module is dust. This remains as a challenging task and many research works are also carried out. In this paper, image processing is used to detect the dust and the percentage of dust is used as a parameter to put into the neural networks. Irradiance depending on the location is taken as one input and the percentage of dust sensor is used to predict the voltage in the panel. If the voltage of the panel is reduced to certain limit then the signal can be triggered to clean the solar panel.