模糊划分线性回归模型在电力负荷预测中的应用

针对电力负荷中长期预测中存在大量的不确定性因素及待预测的负荷变量与关联因素无法很好地满足整个样本序列上预测变量与解释变量间的线性相关性问题,引入了模糊划分理论,构建了基于模糊有序划分的线性回归预测模型。算例应用结果表明,该模型能在较少样本数据基础上实现对电力负荷较为准确的预测,且预测精度较高。

Application of Fuzzy Partition Linear Regression Model to Electric Load Forecasting

There are a lot of uncertain factors within mid-long term power load forecasting. So the linear correlation between the predicted variables and explanatory variables during the entire sample sequence can not be satisfied by the electric load to be predicted and some associated factors. The fuzzy partition theory is introduced to establish linear regression prediction model based optimal order fuzzy partition. An example results show that the proposed method can realize load forecasting with high accuracy in terms of limit sample data.