Amir Nedaei, A. Eskandari, J. Milimonfared, B. Dimd, U. Cali, M. Aghaei
{"title":"A Bi-Level Line-Line Fault Detection Model for Photovoltaic Arrays Using RBM-Based Automatic Feature Extraction","authors":"Amir Nedaei, A. Eskandari, J. Milimonfared, B. Dimd, U. Cali, M. Aghaei","doi":"10.1109/FES57669.2023.10183027","DOIUrl":null,"url":null,"abstract":"Line-Line (LL) faults in PV arrays are usually very difficult to detect due to the production of insufficient fault current under severe conditions such as low mismatch levels and high fault impedances. This paper proposes a very accurate bi-level model to detect and classify LL faults in PV arrays. The model is based on automatic feature extraction using Restricted Boltzmann Machine (RBM) which is respectively combined with a multi-class Support Vector Machine (SVM) classifier and a Random Forest (RF) algorithm in each level. The simulation results show that the proposed model can yield an accuracy of 100% when detecting and classifying various kinds of LL faults.","PeriodicalId":165790,"journal":{"name":"2023 International Conference on Future Energy Solutions (FES)","volume":"67 3 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 International Conference on Future Energy Solutions (FES)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/FES57669.2023.10183027","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Line-Line (LL) faults in PV arrays are usually very difficult to detect due to the production of insufficient fault current under severe conditions such as low mismatch levels and high fault impedances. This paper proposes a very accurate bi-level model to detect and classify LL faults in PV arrays. The model is based on automatic feature extraction using Restricted Boltzmann Machine (RBM) which is respectively combined with a multi-class Support Vector Machine (SVM) classifier and a Random Forest (RF) algorithm in each level. The simulation results show that the proposed model can yield an accuracy of 100% when detecting and classifying various kinds of LL faults.