文章摘要
引用本文:林培杰,陈志聪,吴丽君,程树英.一种PSO-SVM的光伏阵列故障检测与分类[J].福州大学学报(自然科学版),2017,45(5):652~658
一种PSO-SVM的光伏阵列故障检测与分类
Fault detection and classification for photovoltaic arrays based on PSO-SVM
  
DOI:10.7631/issn.1000-2243.2017.05.0652
中文关键词: 光伏阵列  故障  检测  分类  粒子群优化  支持向量机
英文关键词: photovoltaic arrays  fault  detection  classification  particle swarm optimization  support vector machine
基金项目:
作者单位
林培杰 福州大学物理与信息工程学院微纳器件与电池研究所福建 福州 350116 
陈志聪 福州大学物理与信息工程学院微纳器件与电池研究所福建 福州 350116 
吴丽君 福州大学物理与信息工程学院微纳器件与电池研究所福建 福州 350116 
程树英 福州大学物理与信息工程学院微纳器件与电池研究所福建 福州 350116 
摘要点击次数: 163
全文下载次数: 112
中文摘要:
      提出一种粒子群优化支持向量机的光伏阵列故障检测与分类的方法. 分析了光伏阵列输出特性和故障类型,选择合适的特征向量及归一化方式. 选用径向基核函数优化模型结构,并利用PSO算法对参数进行寻优,提高模型精确度. 结合实验平台获取光伏阵列正常工作和8种故障状态的实测数据,随机划分为训练集和测试集,并建立PSO-SVM故障检测与分类模型. 实验表明,应用本模型进行故障检测准确率达99.89%,分类准确率达98.68%,优于BP神经网络以及决策树的检测和分类结果.
英文摘要:
      Fault detection and classification model for photovoltaic arrays is presented by using particle swarm optimization-support vector machine(PSO-SVM). The characteristic and faults of PV arrays are analyzed. Moreover,the appropriate feature vectors are selected and the normalized method is designed,respectively. In order to strengthen the accuracy of the proposed model,the RBF kernel function is applied to improve the model structure,whose parameters are optimized by the PSO algorithm. Based on the measured platform,the experiment data set of the PV array under normal working condition and eight types of faults are recorded. The data set are randomly divided into testing set and training set to train the PSO-SVM model. The accuracy of fault detection and fault classification are 99.89% and 98.68%,respectively,which are superior to those of BP neural network and decision tree.
查看全文   查看/发表评论  下载PDF阅读器
关闭