太原理工大学学报2016，Vol.47Issue(6)：752-757,6.DOI:10.16355/j.cnki.issn1007-9432tyut.2016.06.013
基于相似样本的风速组合预测
Wind Speed Combined Prediction Based on Similar Samples
摘要
Abstract
Wind prediction is an effective means to reduce the adverse effects of large-scale wind power generation on the grid,but the behavior of wind speeds is nonlinear and non-stationa-ry,which yields great challenge for its prediction.This work uses a prediction method for short-term wind speeds,which combines the hierarchical clustering based on grey relation degree algo-rithm (HCGRDA)and simulated annealing fruit fly optimization algorithm based on Gaussian disturbance (GDSAFOA)to optimize the SVM.In this method,the similar sample space is ob-tained by the HCGRDA,and a time series of wind speeds is decomposed by the ensemble empiri-cal mode decomposition (EEMD).The wind speed predication is the linear combination of the SVM based on the chaotic phase space reconstruction model and the dynamic neural network model based on the nonlinear autoregressive models with exogenous inputs (NARX).This meth-od is applied for the model with wind speed data measured from a wind farm in Shanxi.Results show that the proposed method is feasible and competitive.关键词
层次聚类/模拟退火果蝇算法/高斯扰动/灰色关联度/组合预测Key words
hierarchical clustering/simulated annealing fruit fly optimization algorithm/gauss-ian disturbance/grey relation degree/combined prediction分类
信息技术与安全科学引用本文复制引用
谭沛然,马春燕,陈燕,李凌昊,南晓强..基于相似样本的风速组合预测[J].太原理工大学学报,2016,47(6):752-757,6.基金项目
国家电网山西省电力公司资助项目:新能源发电系统对继电保护影响研究与应用(SGSXSZ00FZJS[2015]309) （SGSXSZ00FZJS[2015]309）
