Title:	Advanced Fault Detection and Diagnosis with AI Techniques
Author(s):	M. Chang, J.-L. Li, K.H. Chen, Y.-S. Chen, L. Wang
Keywords:	Fault Detection, Artificial Intelligence, Operation and Maintenance, Failure Mode Diagnosis
Topic:	PV Systems and Storage – Modelling, Design, Operation and Performance
Subtopic:	Operation, Performance and Maintenance of PV Systems
Event:	37th European Photovoltaic Solar Energy Conference and Exhibition
Session:	5DO.4.3
Pages:	1388 - 1391
ISBN:	3-936338-73-6
Paper DOI:	10.4229/EUPVSEC20202020-5DO.4.3
Price:	0,00 EUR
Document(s):	paper, presentation

Abstract/Summary:

---

Operation and maintenance (O&M) is critical for running successful PV projects as the reliable renewable resource for energy transition. We have utilized artificial intelligence (AI) system to optimize O&M task for 120 projects up to 39.2MW more than 2 years. The average daily energy yield of 3.9% (0.13 kWh/kWp) is increased for the projects with AI system, and the O&M transportation cost is 41% reduced. The power prediction algorithm of the fingerprint-like model for each inverter-MPPT is upgraded by simulated POA irradiance instead of that measured by only one pyranometer in multiple-orientation power plants. According to the one-year experiment result for 74 the projects (26.3MW) with 4,792 MPPTs, the Mean Absolute Percentage Error (MAPE) of power prediction decreases from 17.1% to 7.0%. The overall fault detection precision reaches 98.6%, while the precision of failure mode diagnosis improves from 92.3% to 94.0%. The accurate failure mode diagnosis practically enhances operation efficiency and reduces the maintenance expense in the cost structure by low carbon footprint O&M solution.