全球能源互联网2026，Vol.9Issue(1)：45-59,15.DOI:10.19705/j.cnki.issn2096-5125.20240353
基于组合深度学习与核密度估计的架空输电线路载流量概率预测
Probability Prediction of Ampacity for Overhead Transmission Lines Based on Hybrid Deep Learning and Kernel Density Estimation
摘要
Abstract
The ampacity of overhead transmission lines is closely related to meteorological factors such as surrounding wind speed and ambient temperature.Accurate prediction of ampacity is fundamental for fully utilizing transmission line capacity and enhancing the delivery of renewable energy.In this paper,the Complete Ensemble Empirical Mode Decomposition algorithm is employed to decompose meteorological time series into several locally stationary components.Additionally,the Variational Mode Decomposition algorithm is introduced to further process high-frequency components,reducing data complexity.The XGBoost ensemble learning model is used to forecast various meteorological factors,ensuring prediction accuracy and stability.Kernel density estimation is then applied to model the prediction error distribution of meteorological factors.Based on this,a probability prediction method for overhead transmission line ampacity is proposed by integrating deep learning with kernel density estimation.Case studies demonstrate that,compared to two conventional probability prediction methods for ampacity,the proposed approach reduces conservatism indices by 28.54%and 11.93%,respectively,verifying its effectiveness and advantages.关键词
架空输电线路/载流量/深度学习/概率预测/新能源/核密度估计Key words
overhead transmission line/ampacity/deep learning/probabilistic forecasting/renewable energy/kernel density estimation分类
信息技术与安全科学引用本文复制引用
刘洪正,孙成宝,刘建鑫,徐彬,姜在龙,继洋,丁斌..基于组合深度学习与核密度估计的架空输电线路载流量概率预测[J].全球能源互联网,2026,9(1):45-59,15.基金项目
中国南方电网有限责任公司科技项目(GDKJXM20222010). Science and Technology Project of China Southern Power Grid Co.,Ltd.(GDKJXM20222010). （GDKJXM20222010）
