Feature extraction and classification of time-varying power load characteristics based on PCANet and CNN plus Bi-LSTM algorithms

Bian Shaorun; Wang Zhenshu; Song WenhaoZhou Xinhui

首页> 外文期刊>Electric Power Systems Research >Feature extraction and classification of time-varying power load characteristics based on PCANet and CNN plus Bi-LSTM algorithms

【24h】

Feature extraction and classification of time-varying power load characteristics based on PCANet and CNN plus Bi-LSTM algorithms

机译：Feature extraction and classification of time-varying power load characteristics based on PCANet and CNN plus Bi-LSTM algorithms

获取原文

获取原文并翻译 | 示例

掌桥外文数据库（机构版） >>

开具论文收录证明 >>

文献代查 >>

页面导航

摘要
著录项
相关主题

摘要

The feature extraction and classification of power load characteristics are vital for time-varying load modeling. However, due to the influence of the season variation, the existing power load model is difficult to reflect the periodic regularity with seasons. In this paper, a feature extraction and classification method of time-varying power load characteristics based on the Principal Component Analysis Network (PCANet) and CNN+Bi-LSTM algorithm is proposed in order to show the seasonal variation on the power load. By referring to the application of deep learning theory and method in image processing, this paper extracts the features that can reflect the characteristics of seasonal power loads through data mining on daily load curves based on feature extraction. Then, time series encoded daily load curves are transformed into two-dimensional images in polar coordinate system using Gramian Angular Field (GAF) in order to realize machine vision based feature extraction without losing the time-dependence of the data. By two-layer convolution PCA filtering, depth features are extracted using the PCANet algorithm. CNN+Bi-LSTM algorithm is used to perform classification and obtain typical seasonal power load characteristics. The simulation results show that the proposed method is highly accurate and can meet the needs of time-varying load modeling.

著录项

来源
《Electric Power Systems Research》 |2023年第4期|109149.1-109149.10|共10页
作者
Bian Shaorun; Wang Zhenshu; Song WenhaoZhou Xinhui;
展开▼
作者单位

Shandong Univ;

展开▼
收录信息美国《科学引文索引》(SCI);美国《工程索引》(EI);
原文格式 PDF
正文语种英语
中图分类
关键词
Seasonal power load characteristic; Deep learning; Daily load curve; PCANet; CNN plus Bi-LSTM algorithm;

Feature extraction and classification of time-varying power load characteristics based on PCANet and CNN plus Bi-LSTM algorithms

摘要

著录项

相关主题

期刊订阅