Development of a PV modules soiling monitoring system for smart maintenance

This work shows a new method to manage the cleaning of a PV array in a nanogrid. An automated system allows to monitor the soiling level on the front glass of PV modules. In order to quantify the soiling level, we introduce a new sensor concept combining two techniques, namely, optical inspection and I-V curve analysis, that enable not only to quantify the amount of dirt, but also to classify it based on type and covered area. An artificial vision algorithm is specifically developed for this purpose. We created a MATLAB/Simulink model of a DC nanogrid with the aim of studying and smart managing the energy flow among PV array, battery and DC loads. Then, the dust soiling effect has been added to the model to more accurately evaluate the actual power generation. The soiling monitoring system will result in more accurate simulations allowing to develop new control strategies of the nanogrid to optimize the maintenance process. Early tests of the proposed sensor demonstrate its effectiveness.