TITLE:
PV Fault Diagnosis, Including Signal Acquisition, Signal Processing, and Fault Analysis
AUTHORS:
Abdul Bari, Asif Raza Jarwar
KEYWORDS:
PV System, Multi-Fault Situations, Feature Computation, Transient Faults, Artificial Intelligence, Fault Detection and Categorization
JOURNAL NAME:
Journal of Power and Energy Engineering,
Vol.13 No.7,
July
30,
2025
ABSTRACT: The adoption of photovoltaic (PV) systems in modern electrical grids has expanded rapidly due to their economic and environmental benefits. However, these systems are prone to faults—such as partial shading, open circuit, line-to-line, and short circuit—that can significantly reduce energy output and reliability. Timely and accurate fault detection and diagnosis (FDD) are essential for minimizing energy loss, maintenance costs, and system downtime. This paper proposes a Fuzzy Logic Control (FLC)—based approach to detect and classify common DC-side PV faults under dynamic environmental conditions. Three PV configurations—series-parallel (SP), Total Cross-Tied (TCT), and a Hybrid SP-TCT—are modeled in MATLAB/Simulink to simulate fault scenarios. Unlike earlier studies that typically focus on single faults or fail to distinguish between temporary and permanent ones, this research addresses both issues by using a multi-fault detection framework. The FLC model utilizes three key electrical indicators—voltage, current, and power—with threshold values to classify faults as transient or permanent. To ensure robustness, the approach is tested under varying levels of solar irradiance and temperature. Validation is conducted through both simulation and experimental setups using a real PV array. To improve fault localization, wavelet feature extraction and fitness function analysis are incorporated, enhancing the detection of complex fault types such as open circuits and short circuits. Comparative evaluation with existing methods demonstrates the proposed FLC-based system’s superiority in terms of accuracy, adaptability, and real-time capability. The proposed method enables faster and more precise fault classification, supporting improved operational efficiency in PV systems. These findings contribute to the advancement of intelligent FDD systems, particularly in smart grid and large-scale solar applications.