基于偏离度的非等间距灰色预测模型在电力负荷预测中的应用

A Nonequidistance Grey Forecasting Model for Electric Power Based on Deviation Degree Analysis Theory

电力负荷预测过程中,对于原始数据摆动较大,并且数据的整体变化是增大的趋势的序列,以往常采用包络模型来处理它.但是,由于很多不确定的因素的存在,使得上包络及下包络曲线及边缘点的值难以确定,从而削弱了预测值的可信程度.提出了基于偏离度的非等间距灰色预测模型,很好地解决了该类序列的预测问题.并结合实例说明了此模型的可行性和有效性.

In the power load forecast process,The original data for swinging larger,and the tendency of increase overall change,used to adopt envelope model to deal with it.However,because many uncertain factors,makes the envelope curve and the edge points,which is difficult to be determined,the value of the credible degree is weakening,in this paper,Nonequidistance Grey Forecasting Model based on deviation degree analysis was presented,solves the prediction problem such sequence,the feasibility and validity of the model were tested by a case.