Abstract
This paper addresses the challenges resulting from the effects of moisture and rain on the velocity of propagation (VOP) in a spread spectrum time domain reflectometry (SSTDR) monitoring system for detecting and locating faults within a photovoltaic string. A change in the VOP results in shifts of the SSTDR reflection peaks and prevents accurate localization of faults by the monitoring system. Two algorithms, the scale transform and dynamic time warping, are studied as solutions to mitigate these VOP variations. The scale transform algorithm uniformly stretches signals to align with the baseline, while dynamic time warping allows for variable stretching and contracting factors. Both transforms are used to adjust for the changes in VOP to better align data with a known baseline signal, allowing the difference to be used for fault detection and localization. Experimental results demonstrate the influence of rain on VOP within photovoltaic (PV) systems and compare the performance of the scale transform and dynamic time warping algorithms. Both algorithms significantly reduce variability in the data and improve the ability to detect and locate faults within the PV cabling. The scale transform is shown to be generally more robust to the variations.
Funding Information
  • U.S. Department of Energy’s Office of Energy Efficiency and Renewable Energy (EERE) under Solar Energy Technologies Office (DE-EE0008169)

This publication has 28 references indexed in Scilit: