Short-term Wind Power Prediction Based on Feature Selection and Multi-level Deep Transfer Learning

CHENG Kai; PENG Xiaosheng; XU Qiyou; WANG Bo; LIU Chun; CHE Jianfeng

English Version Reader LoginJournal Login

您当前的位置：

首页 >

文章列表页 >

Short-term Wind Power Prediction Based on Feature Selection and Multi-level Deep Transfer Learning

更新时间：2025-12-08

- Short-term Wind Power Prediction Based on Feature Selection and Multi-level Deep Transfer Learning
- Vol. 48, Issue 2, Pages: 497-503(2022)
- 作者机构：
  
  1. 华中科技大学电气与电子工程学院强电磁工程与新技术国家重点实验室
  2. 中国电力科学研究院有限公司新能源与储能运行控制国家重点实验室
- 作者简介：
- 基金信息：
- DOI：
  CLC：
- Published Online：28 February 2022，
  
  Published：2022
- 稿件说明：
移动端阅览
CHENG Kai, PENG Xiaosheng, XU Qiyou, et al. Short-term Wind Power Prediction Based on Feature Selection and Multi-level Deep Transfer Learning[J]. 2022, 48(2): 497-503.
DOI：

CHENG Kai, PENG Xiaosheng, XU Qiyou, et al. Short-term Wind Power Prediction Based on Feature Selection and Multi-level Deep Transfer Learning[J]. 2022, 48(2): 497-503. DOI：

摘要

风电功率预测对电力系统的安全运行与经济调度至关重要，但对新建、扩容或改造的风电场功率预测建模面临两大难题：首先，新建场站及场站扩容造成部分场站运行数据不足，模型训练不充分；其次，传统浅层神经网络难以应对愈发复杂的预测输入信息。为此，提出了一种基于特征选择多层级深度迁移学习的风电场短期功率预测模型。首先，采用相关性分析方法对同省其他风电场的历史数据集按与目标风电场的相关性划分层级，然后按相关性由低到高的顺序，将源风电场预测模型迁移到目标风电场，最后采用特征选择方法优化迁移模型，保证相关性强的特征有效迁移。算例分析表明：1）多层级深度迁移学习模型可以弥补新建风电场训练样本不足的难题，与直接建模相比，精度提升6.5%;2）采用特征选择方法优化之后的模型，预测精度可提升0.4%，因而所提出的方法是数据短缺情况下一种有效的风电场功率预测建模方法。

Abstract

Wind power prediction is very important to the safe operation and economic dispatch of the power system

but the power prediction modeling of newly built

expanded or renovated wind farms faces two major problems: 1) New stations and station expansions cause some stations to operate insufficient data

insufficient model training. 2) Traditional shallow neural networks are difficult to cope with increasingly complex prediction input information. Therefore

a short-term wind power prediction model based on feature selection multi-level deep transfer learning is proposed in this paper. First

correlation analysis method to classify the historical data sets of other wind farms in the same province according to the correlation with the target wind farm is proposed. Then

the source wind farm prediction model to the target wind farm in the order of the correlation from low to high is migrated. Finally

the feature selection method is adopted to optimize the migration model and to ensure the effective migration of relevant features. The analysis of calculation examples shows that: 1) The multi-level deep transfer learning model can make up for the problem of insufficient training samples for new wind farms. Compared with direct modeling

the accuracy is improved by 6.5%. 2) The model optimized by the feature selection method can increase the predicted accuracy by 0.4%

so the method proposed in this paper is an effective method for power prediction of wind farms with data shortage.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Short-term Wind Power Prediction Based on Dynamic Cluster Division and BLSTM Deep Learning Method

Short-term Power Prediction of Wind Power Cluster Based on SDAE Deep Learning and Multiple Integration

功率模块多物理场快速求解技术综述

Robust Allocation of Energy Storage System for Multiple Wind Farms

Uncertainty Evaluation of Wind Power Prediction Based on Monte-Carlo Method

Related Author

YANG Zimin

LANG Jianxun

WANG Hongyu

LI Cong

PENG Xiaosheng

WANG Haohuai

CHE Jianfeng

WANG Bo

Related Institution

State Key Laboratory of New Energy and Storage Operation Control, China Electric Power Research Institute

Power Dispatching Control Center of China Southern Power Grid

State Key Laboratory of Strong Electromagnetic Engineering and New Technology, School of Electrical and Electronic Engineering, Huazhong University of Science and Technology

Center for Power Electronics Systems (CPES), Virginia Tech, Blacksburg

State Key Laboratory of Electrical Insulation and Power Equipment(Xi'an Jiaotong University)

AI问答

⁰