| 应用于光伏阵列故障诊断的DA-SOM算法研究 |
| |
| 引用本文： | 张晓阳,李田泽,张涵瑞,韩鸿雁,李兵.应用于光伏阵列故障诊断的DA-SOM算法研究[J].电源学报,2022,20(2):122-128. |
| |
| 作者姓名： | 张晓阳 李田泽 张涵瑞 韩鸿雁 李兵 |
| |
| 作者单位： | 山东理工大学电气与电子工程学院,山东理工大学电气与电子工程学院,山东理工大学电气与电子工程学院,山东理工大学电气与电子工程学院,国网德州供电公司 |
| |
| 基金项目： | 山东省研究生教育资助项目（基金编号：SDYKC18067） |
| |
| 摘 要： | 针对自组织映射SOM(self-organizing map)神经网络聚类性能易受其初始权值的影响,采用蜻蜓算法DA(dragonfly algorithm)优化SOM神经网络的局部权重失真指数LWDI(locally weighted distortion index),对神经网络的初始权值进行寻优.以光伏阵列故障数...
|
| 关 键 词： | 自组织映射神经网络 蜻蜓算法 光伏阵列 故障诊断 仿真 |
| 收稿时间： | 2019/11/8 0:00:00 |
| 修稿时间： | 2022/4/11 0:00:00 |
Study on DA-SOM Algorithm Applied to Fault Diagnosis of Photovoltaic Array |
| |
| ZHANG Xiaoyang,LI Tianze,ZHANG Hanrui,HAN Hongyan,LI Bing.Study on DA-SOM Algorithm Applied to Fault Diagnosis of Photovoltaic Array[J].Journal of power supply,2022,20(2):122-128. |
| |
| Authors: | ZHANG Xiaoyang LI Tianze ZHANG Hanrui HAN Hongyan LI Bing |
| |
| Affiliation: | Shandong University of Technology,Shandong University of Technology ,College of Electrical and Electronic Engineering,,, |
| |
| Abstract: | In view of the clustering performance of self-organizing map(SOM) neural network which is susceptible to initial weights, the dragonfly algorithm(DA) was used to optimize the locally weighted distortion index(LWDI), thus op-timizing the initial weights of the SOM neural network. The photovoltaic(PV) array failure data samples were taken as the research object in this paper. By comparing and analyzing the output characteristics under normal and failure con-ditions, a fault diagnosis model was established and further used to diagnose and output the fault types. Simulation re-sults show that compared with the basic SOM, back propagation(BP) and DA-BP neural networks, the DA-SOM neural network can obtain better clustering results and improve the accuracy of PV array fault diagnosis. |
| |
| Keywords: | SOM dragonfly algorithm photovoltaic array fault diagnosis the simulation |
|
| 点击此处可从《电源学报》浏览原始摘要信息 |
|
点击此处可从《电源学报》下载免费的PDF全文 |
|
