Solar Panels String Predictive and Parametric Fault Diagnosis Using Low-Cost Sensors

Area of Science:

• Renewable Energy Systems
• Electrical Engineering
• Photovoltaic Technology

Background:

• Solar panel degradation can lead to significant energy loss and system failure.
• Existing monitoring systems may lack real-time predictive capabilities for early fault detection.
• Effective fault diagnosis is crucial for maintaining the performance and longevity of photovoltaic installations.

Purpose of the Study:

• To propose a real-time supervision and predictive fault diagnosis method for solar panel strings.
• To enable early detection and isolation of fault symptoms using Voc-Isc curve analysis.
• To develop a scalable and integrable architecture for conventional photovoltaic systems.

Main Methods:

• Analysis of Voc-Isc curves for fault symptom detection and parametric isolation.
• High-frequency data acquisition using ESP8266 modules and low-cost sensors (ASC712-5A, FZ0430).
• Data transmission via internet to a SCADA system (iFIX V6.5) using Modbus TCP/IP and OPC protocols.
• Experimental determination of detection thresholds via inductive shading.

Main Results:

• The method provides systematic, online, automatic, and permanent predictive supervision.
• It allows for early detection and isolation of fault symptoms with a sufficient margin for intervention.
• The architecture is designed to be scalable and integrable into existing photovoltaic installations.
• The use of low-cost technology makes the solution economically viable.

Conclusions:

• The proposed method offers effective real-time predictive fault diagnosis for solar panel strings.
• Early detection and automatic disconnection prevent cumulative degradation and irrecoverable failures.
• The system's scalability and low-cost components facilitate widespread adoption in photovoltaic installations.