A NOVEL METHOD FOR OPTICALLY MONITORING PV TRACKING SYSTEMS INCLINATION USING WAVELET TRANSFORMS AND NEURAL NETWORKS

Abstract

This work demonstrates the feasibility of utilizing optical images of PV modules or arrays controlled by tracking systems to determine its tilt angle. The proposed novel method uses discrete wavelet transforms and artificial neural networks. The process can be accomplished remotely to reduce the need of manpower and time required to find the tracking systems inclination angle. Hardware and software needed is relatively simple. Application of this method can be extended to other systems such as solar thermal collectors. Proposed algorithms performance, using different artificial neural network parameters and families of wavelet transforms, is also investigated. All combinations illustrate the reliability and practicality of the proposed method. Moreover, in order to account for noisy images, the performance is examined under different values of signal to noise ratios, the method performed very well at all noise levels introduced.