摘要
提出一种基于数据挖掘技术的电力负荷短期预测方法,将SVM方法引入到短期负荷预测研究领域。通过随机选取历史负荷数据来更新回归函数,这样可以充分保证计算速度和较高的预测精度。提出利用松原地区的历史负荷数据作为训练样本,通过与传统的BP神经网络预测模型进行对比,对预测结果进行比较,证明SVM预测方法在一定程度上能够保证短期负荷预测的精度。
This paper presented a short-term load forecast method that was based on the data mining technique and introduced the SVM method into the field of short-term load forecast. By randomly selecting historical load data to update the regression function,the computing speed and higher forecasting accuracy could be fully guaranteed. The paper used the historical load data of Songyuan areas as training samples, compared them with the traditional BP neural network prediction models, analyzed the predicted results and finally proved that the SVM forecasting method could ensure the accuracy of the short-time load forecasting to some degree.
出处
《电测与仪表》
北大核心
2014年第18期1-4,共4页
Electrical Measurement & Instrumentation
关键词
电力系统
气象因素
支持向量机
短期负荷预测
power system, meteorological factor, support vector machines, short-term load forecast
