Fault Diagnosis in a Grid-Connected Photovoltaic Systems Based on Hierarchical Clustering

EasyChair Preprint no. 6533

Abstract

This paper proposes an effective fault detection and diagnosis (FDD) of Grid-Connected Photovoltaic (GCPV) systems. The developed approach combines the advantages of both Principal Component Analysis (PCA) model and Hierarchical Clustering (HC) scheme. The PCA model is applied to extract and select the most informative features from GCPV system data. While, the HC metric is used to classify the GCPV faults and distinguish between the operating healthy and faulty modes. The proposed FDD approach, the socalled PCA-based HC is experimentally tested and validated using GCPV system data. Different case studies are investigated in this paper in order to illustrate the efficiency and the robustness of the proposed framework. A comparison with well-known techniques is also presented. The obtained results confirm the high accuracy of the developed technique.

Keyphrases: fault classification, fault diagnosis, feature extraction and selection (FES), Grid-connected PV systems, Hierarchical clustering (HC), Principal Component Analysis (PCA)

@Booklet{EasyChair:6533,
  author = {Amal Hichri and Mansour Hajji and Majdi Mansouri and Hazem Nounou and Abdelmalek Kouadri and Kais Bouzrara},
  title = {Fault Diagnosis in a Grid-Connected Photovoltaic Systems Based on Hierarchical Clustering},
  howpublished = {EasyChair Preprint no. 6533},

  year = {EasyChair, 2021}}