Assessing dust on PV modules using image processing techniques

Abstract

Accumulation of dust on PV modules is one of the major issues limiting the technology adoption in many regions of the world, especially the in Middle East. Even though the solar irradiation is relatively high, the dusty environment is one of the key factors limiting a larger scale implementation of such renewable energy sources. In this paper we present a new approach for optimizing operation of a PV plant in terms of dust assessment and management. The proposed method includes analysis of PV modules' surfaces using an image processing technique. The acquisition of images is assumed to be performed by aerial robotics (i.e. drones) equipped with high resolution cameras. The key novelty of this work is the algorithm used for dust detection and the method used for assessing the level of dust on a given surface. This is done by using MATLAB image processing tool box. In this paper, the concept and methodology are presented where the algorithm is validated against electrical measurement where it showed a variation of 5% to 5.5% of detection variation based on the method used in the detection algorithm.