Online reduced kernel GLRT technique for improved fault detection in photovoltaic systems. (15th July 2019)

• Record Type: Journal Article
• Authors: Fezai, R.; Mansouri, M.; Trabelsi, M.; Hajji, M.; Nounou, H.; Nounou, M.
• Is Part Of: Energy. Volume 179(2019)
• Page Start: 1133
• Page End: 1154
• Publication Date: 2019-07-15
• Subjects: Fault detection -- Photovoltaic (PV) system -- Kernel principal component analysis (KPCA) -- Kernel generalized likelihood ratio test (KGLRT) -- Online reduced GLRT (OR-GLRT); Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2019.05.029
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: This paper proposes an effective kernel generalized likelihood ratio test (KGLRT) technique for fault detection in Photovoltaic (PV) systems. The proposed technique is considered as an improvement of the conventional KGLRT with extended online capabilities and lower computational complexity. The proposed online reduced KGLRT (OR-KGLRT) is based on transforming the process data into a higher dimensional space (where the data becomes linear), which makes the kernel-based scheme attractive for modeling nonlinear systems. The performance of the proposed method is evaluated and compared to the conventional KGLRT statistic using a simulated PV data. Both techniques are applied to detect single and multiple failures (including Bypass, Mismatch, Mix and Shading failures). The selected performance criteria are the good detection rate (GDR), false alarm rate (FAR), and computation time (CT). Simulation results show superior detection efficiency of the proposed approach compared to the conventional KGLRT statistic in terms of GDR, FAR and CT.