...
首页> 外文期刊>Renewable energy >Multiclass adaptive neuro-fuzzy classifier and feature selection techniques for photovoltaic array fault detection and classification
【24h】

Multiclass adaptive neuro-fuzzy classifier and feature selection techniques for photovoltaic array fault detection and classification

机译:用于光伏阵列故障检测和分类的多类自适应神经模糊分类器和特征选择技术

获取原文
获取原文并翻译 | 示例
           

摘要

In this paper, a Multiclass Adaptive Neuro-Fuzzy Classifier (MC-NFC) for fault detection and classification in photovoltaic (PV) array has been developed. Firstly, to show the generalization capability in the automatic faults classification of a PV array (PVA), Fuzzy Logic (FL) classifiers have been built based on experimental datasets. Subsequently, a novel classification system based on Adaptive Neuro-fuzzy Inference System (ANFIS) has been proposed to improve the generalization performance of the FL classifiers. The experiments have been conducted on the basis of collected data from a PVA to classify five kinds of faults. Results showed the advantages of using the fuzzy approach with reduced features over using the entire original chosen features. Then, the designed MC-NFC has been compared with an Artificial Neural Networks (ANN) classifier. Results demonstrated the superiority of the MC-NFC over the ANN-classifier and suggest that further improvements in terms of classification accuracy can be achieved by the proposed classification algorithm; furthermore faults can be also considered for discrimination. (C) 2018 Elsevier Ltd. All rights reserved.
机译:本文开发了一种用于光伏阵列中故障检测和分类的多分类自适应神经模糊分类器(MC-NFC)。首先,为了展示光伏阵列(PVA)自动故障分类中的泛化能力,基于实验数据集建立了模糊逻辑(FL)分类器。随后,提出了一种基于自适应神经模糊推理系统(ANFIS)的新型分类系统,以提高FL分类器的泛化性能。实验是基于从PVA收集的数据进行的,以对五种故障进行分类。结果表明,与使用整个原始选定特征相比,使用具有减少特征的模糊方法的优势。然后,将设计的MC-NFC与人工神经网络(ANN)分类器进行了比较。结果证明了MC-NFC优于ANN分类器的优势,并表明通过提出的分类算法可以实现分类精度的进一步提高。此外,也可以考虑将故障用于区分。 (C)2018 Elsevier Ltd.保留所有权利。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号