Assisting PV Experts in On-Site Condition Evaluation of PV Modules UsingWeather-Independent Dark IV String Curves, Artificial Intelligence and aWeb-Database

被引：0

作者：

Rueter, Joachim ^{[1
]}

Meyer, Felix ^{[1
]}

Behrens, Grit ^{[1
]}

Mertens, Konrad ^{[2
]}

Diehl, Matthias ^{[3
]}

机构：

[1] Univ Appl Sci Bielefeld, Solar Comp Lab, Artilleriestr 9, D-32425 Minden, Germany

[2] Munster Univ Appl Sci, Testing Lab Photovolta, Elect Engn, Stegerwaldstr 39, D-48565 Steinfurt, Germany

[3] Photovoltaikburo Ternius & Diehl GbR, Schunauerhofstrstr 27, D-65428 Russelsheim, Germany

来源：

ADVANCES AND NEW TRENDS IN ENVIRONMENTAL INFORMATICS: A BOGEYMAN OR SAVIOUR FOR THE UN SUSTAINABILITY GOALS? | 2022年

关键词：

Artificial intelligence; Deep learning; Photovoltaic; Dark IV string curve; IV curve; Maximum power point; Web; Database; Application; Predictive maintenance;

D O I：

10.1007/978-3-030-88063-7_6

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

Photovoltaic (PV) modules can make a huge contribution to achieve the Sustainable Development Goals of the United Nations. To be able to make that contribution, regular check-ups and evaluation of installed PVmodules are necessary as they can develop faults and degenerate over time. In this project, we improve the dark IV string curve method used for on-site fault detection and module evaluation. We do so by training artificial intelligence (AI) models to predict the maximum power point and the bright IV curve of PV modules given the weather-independent dark IV string curve. We present some background on this topic, describe the data used for training and the developed models. The results are illustrated graphically. To make the models available for PV experts in practice and to support their decisionmaking process, we also developed the web-database-application iPVModule for storing historical PV Module data and integrated the AI-models.

引用

页码：87 / 102

页数：16