基于小波分析法与滚动式时间序列法的风电场风速短期预测优化算法

来源期刊:中南大学学报(自然科学版)2010年第1期

论文作者:刘辉 田红旗 李燕飞

文章页码:370 - 375

关键词:风速预测;滚动式时间序列法;小波分析法;时间序列分析法;优化算法

Key words:wind speed forecast; rolling time series method; wavelet analysis method; time series method; optimization algorithm

摘    要:为实现风电场风速的超前多步高精度预测,提出一种基于小波分析法与滚动式时间序列法混合建模的优化算法。该优化算法引入小波分析法对风电场实测非平稳风速序列进行分解重构计算,将非平稳性原始风速序列转化为多层较平稳分解风速序列,利用对传统时间序列分析法改进后的滚动式时间序列法对各分解层风速序列建立非平稳时序预测模型,并通过模型方程实现超前多步滚动式预测计算。仿真结果表明:该优化算法实现了风速的高精度短期多步预测,将传统时间序列分析法对应超前1步、3步、5步的预测精度分别提高了54.22%,26.44%和19.38%,其预测的平均相对误差分别为1.14%,3.06%和4.41%;优化算法具有较强的细分与自学习能力。

Abstract: In order to achieve high-precision multi-step ahead prediction for real-time wind speed data that sampled from wind farm, based on wavelet analysis method and rolling time series method, a forecast improved algorithm was proposed. Wavelet analysis method was used to make decomposition and reconstruction calculations for original wind speed series, and multi-layer more steady wind speed series was obtained. Then rolling time series method that was modified from traditional time series method was used to build unsteady prediction models for each layer, and corresponding equations were used to realize multi-step rolling forecast calculation. Simulation results show that the optimization algorithm attains high-precision multi-step ahead forecast results, improves forecast precision of one-step, three-step, five-step ahead forecast traditional time series method by 40.48%, 29.22%, 45.73%, respectively, and the mean relative error is only 1.72%, 3.61%, 7.12%, respectively. The optimization algorithm has respectively excellent subdivision and self-learning ability.

基金信息:国家“十一五”科技支撑计划重点项目
国家留学基金资助项目
中南大学优秀博士学位论文扶植基金资助项目

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 主办 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(湘)字第028号   湘ICP备09001153号