基于马尔可夫链筛选组合预测模型的中长期负荷预测方法

摘要

在负荷预测的模型组合过程中，主要是根据历史数据的趋势恰当选择模型，再根据模型特点选择权重分配方法。针对灰色关联度满足要求的几种模型预测值分化较大的问题，从负荷数据的增长率无后效性这一特点出发，通过对原始数据增长率的分析，采用马尔可夫链划分区间，从几种满足精度要求的模型中筛选出两种进行组合预测，通过方差—协方差方法分配权重。经过该种方法的筛选，不仅可以更准确地选择组合预测模型的类型，而且具有较高精度。%It is important to choose theright model according to thetrend of the historical data in the process of load forecast model combination.And then, a method is chosen to assign weights according to thefeaturesof the models.Even forecast models meetthe requirements of the grey correlation degree, theforecast results stillhave large differences. To solve the question, this paper, according to the feature that the growth rate of load data isnon-aftereffect property of Markov chain, and byanalyzingthegrowth rate of load data, usesMarkov chain to divideintervals andscreens two kinds from the models which havemet theaccuracyrequirement, and adopts the method ofvariance-covariance toassign weights. Using this method of screening not only canaccurately choose the models forcombination forecast, but also has a high precision.