Events and Trade Shows

Review of AI Methods for Fault Diagnosis and Predictive Maintenance in Solar Panels

Hannes Peter Wasser and Patrali Majumder, Fraunhofer IFF

At this year's CPE-POWERENG, Hannes Peter Wasser and Patrali Majumder will present their latest paper on the role of artificial intelligence in anomaly detection and predictive maintenance of photovoltaic systems.

The paper offers:

• a comprehensive analysis of typical malfunctions in PV systems, their probability of occurrence and impact on energy yield,
• a detailed overview of the status of AI-based diagnostic systems in the context of PV maintenance,
• an evaluation of different AI methods—including machine learning, deep learning and neural networks—with regard to their practical suitability for detecting and classifying PV malfunctions,
• as well as an assessment of the transferability of these methods to real PV systems, taking into account electrical and environmental parameters.

The Fraunhofer IFF thus participates in the international discourse on increasing the efficiency of solar plants using digital technologies and demonstrates how AI solutions can be implemented in practice.