基于预测误差校正的支持向量机短期风速预测  

Short-term Wind Speed prediction with Support Vector Machine Based on Predict Error Correction

作　　者：周松林[1,2] 茆美琴[1] 苏建徽[1] 

机构地区：[1]合肥工业大学教育部光伏系统工程中心,合肥230009 [2]铜陵学院电气工程系,铜陵244000

出　　处：《系统仿真学报》2012年第4期769-773,共5页Journal of System Simulation

基　　金：国家重点基础研究发展计划(973计划)(2009CB219708)

摘　　要：对风电场风速进行较为准确的预测,对于电力部门及时调整调度计划至关重要。建立了支持向量机风速预测模型,并提出了结合预测误差校正来提高预测精度的新思路。先建立SVM模型初步预测风速,再将得到的训练误差和测试误差分别构建样本,建立基于小波-支持向量机的误差预测模型进行误差预测,最后用预测误差对风速初步预测值进行校正。仿真结果表明所提方法能明显改善预测精度,而且方法简洁明了,具有很好的稳健性,能够推广到长期风速预测、负荷预测及其它预测领域。An accurate predict of wind speed is important for power department to regulate dispatching plan in time.A support vector machine（SVM） model was established for forecasting wind speed,at the same time,a new idea using predict error correction methods to improving the prediction accuracy was proposed.SVM model was established for a preliminary prediction of wind speed,and then the error prediction model based on wavelet-support vector machine was set up by use of samples which were separately constructed from training error and testing error.Finally,the correction of preliminary prediction values was carried out.Simulation results show that the proposed method can significantly improve the prediction accuracy,and the method is simple,clear and steady,which can be extended to long-term wind speed prediction,load prediction,and other prediction field.

关 键 词：风速预测 预测误差校正 支持向量机 小波分解 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]