Artificial intelligence and internet of things to improve efficacy of diagnosis and remote sensing of solar photovoltaic systems: Challenges, recommendations and future directions. (June 2021)

• Record Type: Journal Article
• Authors: Mellit, Adel; Kalogirou, Soteris
• Is Part Of: Renewable & sustainable energy reviews. Volume 143(2021)
• Publication Date: 2021-06
• Subjects: Deep learning -- Fault detection and diagnosis -- Internet of things -- Machine learning -- Photovoltaic systems -- Remote sensing -- Smart monitoring; Renewable energy sources -- Periodicals; Power resources -- Periodicals; Énergies renouvelables -- Périodiques; Ressources énergétiques -- Périodiques; 333.794
• DOI: 10.1016/j.rser.2021.110889
• Languages: English
• ISSNs: 1364-0321

Abstract

Abstract: Currently, a huge number of photovoltaic plants have been installed worldwide and these plants should be carefully protected and supervised continually in order to be safe and reliable during their working lifetime. Photovoltaic plants are subject to different types of faults and failures, while available fault detection equipment are mainly used to protect and isolate the photovoltaic plants from some faults (such as arc fault, line-to-line, line-to-ground and ground faults). Although a good number of international standards (IEC, NEC, and UL) exists, undetectable faults continue to create serious problems in photovoltaic plants. Thus, designing smart equipment, including artificial intelligence and internet of things for remote sensing and fault detection and diagnosis of photovoltaic plants, will considerably solve the shortcomings of existing methods and commercialized equipment. This paper presents an overview of artificial intelligence and internet of things applications in photovoltaic plants. This research presents also the most advanced algorithms such as machine and deep learning, in terms of cost implementation, complexity, accuracy, software suitability, and feasibility of real-time applications. The embedding of artificial intelligence and internet of things techniques for fault detection and diagnosis into simple hardware, such as low-cost chips, may be economical and technically feasible for photovoltaic plants located in remote areas, with costly and challenging accessibility for maintenance. Challenging issues, recommendations, and trends of these techniques will also be presented in this paper. Highlights: Remote sensing of PV systems based on IoT technique is discussed. Application of machine learning and deep learning in FDD of PV systems are presented. A smart configuration based on IoT technique and embedded intelligent FDD is suggested. Challenges, recommendations and future directions in this area are presented.