摘要
介绍了一种基于自适应线性逻辑网络(ALLN)的风电功率预测方法,描述了ALLN算法的特点、基本结构、建模方法以及预测与学习方法;采用该方法进行了风速预测试验,对试验结果进行了对比分析,详细分析了影响试验结果的因素。首次提出了模型适应性的概念,并给出了计算公式。从模型训练输入数据和预测输入数据的角度,分析了影响ALLN算法模型适应性与预测结果的主要因素。
A wind power forecasting method based on adaptive linear logic network （ALLN） is introduced in this paper. The characteristics, basic mechanism and modeling methods, prediction and studying methods of ALLN algorithm are given in details. Wind speed forecasting experiments are carried out and then experimental results are compared, analyzed and evaluated to obtain factors which effect on testing results. The concept of model adaptability and its definition are brought forward for the first time. In the end, the main factors which have effects on ALLN forecasting results are analyzed viewing from input data used for model training and input data used for wind speed forecasting.
出处
《可再生能源》
CAS
北大核心
2013年第6期61-65,共5页
Renewable Energy Resources
基金
中国华电集团科技项目(CHFKJ12-01-22)
关键词
风电功率预测
短期预测
自适应线性逻辑网络
预测性能评估
预测模型
wind power forecast
short-term forecast
Adaptive Linear Logic Network
assessment for prediction performance
prediction model
