基于LDA-IBES-RELM的光伏阵列故障诊断方法
Fault diagnosis method of PV array based on LDA-IBES-RELM
邹凯 1曾宪文 2王洋 1高桂革1
作者信息
- 1. 上海电机学院 电气学院, 上海 201306
- 2. 上海电机学院 电子信息学院, 上海 201306
- 折叠
摘要
针对光伏阵列故障诊断准确率偏低的问题,提出了一种基于改进秃鹰搜索算法(IBES)优化正则化极限学习机(RELM)的故障诊断方法.首先在Simulink建立光伏阵列仿真模型,模拟典型故障并提取故障特征数据,同时利用线性判别分析(LDA)对特征量降维作为故障诊断模型的输入;其次利用Logistic混沌映射、Levy飞行策略和柯西高斯变异扰动策略对秃鹰算法进行改进;最后将IBES用于对RELM的隐层参数寻优.实验结果表明:LDA-IBES-RELM模型与BES-RELM、IBES-RELM模型对比,得到的故障诊断准确率为97.71%,优于其他两种模型,验证了LDA-IBES-RELM模型用于光伏阵列故障诊断的有效性和实用性.
Abstract
In response to the problem of low accuracy in fault diagnosis of photovoltaic arrays,this paper proposes a fault diagnosis method based on the improved bald eagle search(IBES)optimized regularized extreme learning machine(RELM).The paper first establishes a photovoltaic array simulation model in Simulink,simulates typical faults and extracts fault feature data.Then,linear discriminant analysis(LDA)is used to reduce the dimensionality of the feature quantity as input for the fault diagnosis model.Secondly,logistic chaotic mapping,Levy flight strategy,and Cauchy Gaussian mutation perturbation strategy are used to improve the bald eagle algorithm.Finally,IBES is used to optimize the hidden layer parameters of RELM.Finally,the LDA-IBES-RELM model proposed in the article was compared with the BES-RELM and IBES-RELM models,and the fault diagnosis accuracy of the proposed model was 97.71%,which was superior to the other two models.This verified the effectiveness and practicality of the LDA-IBES-RELM model for PV array fault diagnosis.
关键词
正则化极限学习机/光伏阵列/故障诊断/改进秃鹰搜索算法/线性判别分析Key words
regularized extreme learning machine(RELM)/PV array/fault diagnosis/improved bald eagle search(IBES)/linear discriminant analysis(LDA)引用本文复制引用
基金项目
国家自然科学基金青年基金资助项目(62003206)
出版年
2024
NETL
NSTL
万方数据