基于优化决策树的短期电力负荷预测算法

考虑非负荷因素对短期负荷预测的影响,提出了基于优化决策树的短期负荷预测新方法。该方法先对样本数据进行预处理,利用粗糙集理论对决策树的测试属性约简;然后针对决策树ID3算法的缺陷,改进测试属性选择的方法,提出优化算法M ID3。通过2次优化建立的决策树短期负荷预测模型,在有更好的分类准确率的前提下具有较小的规模。结果表明,该方法可提高短期负荷预测的精度,具有一定的实用性和优势。

Short-term load forecasting based on optimized decision tree algorithm

Considering the impact coming from non-load factors,this paper introduces a new short-term load forecasting method based on optimized decision tree.This method first preprocess the sample data,reduces the testing attributes of decision tree by using rough set,then it presents an optimized MID3 algorithm by improving the selection of testing attributes. After twice optimization,the established short-term load forecasting model has smaller scale with better categorizing accuracy.Testing results show that this method can improve the accuracy of short-term load forecasting.