基于业扩报装的月度负荷预测

A Monthly Load Forecasting Method Based on Business Expansion and Installation

随着我国经济进入新常态,产业结构调整加大了电力中短期负荷预测难度。采用支持向量机法、业扩增量调整法、温度调整法和K-L信息量法,充分考虑温度、经济等主导因素的影响,使用2007—2015年业扩报装数据,建立训练样本,利用历史数据检验模型的效性,并与其他方法进行比较,剖析了2016年下半年浙江省月度负荷。结果表明所选方法选取合理,预测结果相对较优。

With China′ s economy entering a new normal, industrial restructuring makes medium and shortterm load forecasting more difficult. The paper adopts support vector machine method, business expansion and installation adjustment method, temperature regulation method, K-L information and completely takes important influencing factors including temperature and economy into account; besides, it uses business expansion and installation data from 2007 to 2015 to acquire training sample and compares with other methods to analyze monthly load in Zhejiang province in the second half of 2016. The result implies that the methods are reasonable and superior in prediction.