几种短期风速时间序列预测技术的比较  

A Comparison of Various Forecasting Techniques Applied to Mean Hourly Wind Speed Time Series

作　　者：杨小强[1] 韩亚军[1] 刘静[1] 

机构地区：[1]重庆科创职业学院机电工程学院,重庆402160

出　　处：《常州大学学报（自然科学版）》2016年第1期88-92,共5页Journal of Changzhou University：Natural Science Edition

基　　金：重庆市第二批高等学校青年骨干教师资助计划(渝教人〔2013〕74号)

摘　　要：对风速进行准确的预测可以减轻对电力系统的不利影响,提高风电场在电力市场中的竞争力。比较了几种不同的风速预测方法,它们都是采用时间序列分析短期风速数据。讨论传统的线性自回归滑动平均模型(ARMA),常用的前馈和循环神经网络,同时对自适应神经模糊推理系统(ANFIS)以及神经逻辑网络进行比较。通过建模对几种方法的预测性能进行估计,最终得出基于人工智能的模型比线性模型效果更好,能够准确快速地预测结果。Accurate prediction wind speed can reduce the iMPact on the power system and improve the com- petitiveness of the wind farm in the electric power market. This paper mainly compared several different wind speed forecasting methods, which use the time series to analyze the short-term wind speed data,dis- cussed the traditional linear auto regressive moving average model（ARMA）, the common feed forward and recurrent neural network,and also compared the adaptive neural fuzzy inference system（ANFIS） and the neural logic network.The model shows that the artificial intelligence model is better than the linear model, and it can predict the wind speed accurately and quickly.

关 键 词：风速 时间序列 神经网络 

分 类 号：TM614[电气工程—电力系统及自动化]