Fault Detection and Diagnosis of PV Systems Using Kalman-Filter Algorithm Based on Multi-zone Polynomial Regression
Langue
EN
Chapitre d'ouvrage
Ce document a été publié dans
Recent Developments in Model-Based and Data-Driven Methods for Advanced Control and Diagnosis. ACD 2022. Studies in Systems, Decision and Control. 2023-06-14
Résumé en anglais
The correct functioning of the PV system with the desired operational efficiency, under all circumstances, has become a strategic research challenge. Toward this end, faults must be timely diagnosed, detected, and identified ...Lire la suite >
The correct functioning of the PV system with the desired operational efficiency, under all circumstances, has become a strategic research challenge. Toward this end, faults must be timely diagnosed, detected, and identified to enhance reliability, maintainability, and safety of the PV system. In this context, this paper presents a novel FDD methodology based on Kalman filter (KF) through a model-based approach for online health monitoring of the PV system’s DC side. This approach is formulated on the comparison of two different regression analysis techniques of PV characteristics under global max power point (GMPP) at array level. The first is based on a non-linear polynomial regression equation, whereas the second is based on a multivariate regression method. In particular, the proposed methods show their effectiveness to detected anomalies at low-current with sensor-less irradiation. The performance of the proposed diagnosis methods is evaluated under Matlab/Simulink® with varying environmental conditions.< Réduire
Mots clés en anglais
PV system
Monitoring
Kalman filter
Non-linear polynomial regression
Multivariate polynomial regression
Fault diagnosis and detection
Unités de recherche