Predictive Fault Diagnosis for Ship Photovoltaic Modules Systems Applications

Area of Science:

• Marine Engineering
• Renewable Energy Systems
• Predictive Maintenance

Background:

• Solar power generation systems are crucial for renewable energy supply on ships.
• Photovoltaic modules (PVMs) on ships face harsh conditions, leading to accelerated aging and degradation.
• Common failure modes like hot-spots pose significant risks to PVMs.

Purpose of the Study:

• To develop an application for managing and supervising solar power generation systems using predictive fault diagnosis (PFD).
• To apply PFD techniques for early detection and isolation of faults in shipboard photovoltaic modules.
• To mitigate cumulative degradation and prevent severe failure modes through timely maintenance interventions.

Main Methods:

• Utilized a National Marine Electronics Association (NMEA) 2000 smart sensor network for data acquisition.
• Applied online analysis of time-series data for current-voltage (I-V) parameters.
• Employed comparative trend analysis of power generation data, using decreased power as a predictor symptom parameter (PS).

Main Results:

• The developed PFD method allows for early fault detection and isolation in shipboard solar power systems.
• Identified decreased generated power as an effective predictor symptom parameter for PVM degradation.
• Demonstrated that comparative trend analysis of NMEA sensor data can diagnose failure modes like hot-spots.

Conclusions:

• PFD, integrated with NMEA 2000 sensors, provides effective early fault detection for shipboard solar power systems.
• This approach enables proactive maintenance, preventing the progression of degradation and critical failures.
• Comparative trend analysis of power parameters is a reliable method for diagnosing PVM health in marine environments.