具有白指数律重合性的改进GM(1,1)负荷预测

针对传统灰色模型的建模机理和存在的局限性,提出了改进方法。首先对原始负荷数据序列应用对数处理的数据预处理方法,减小了序列的级比偏差,然后将具有白指数律重合性的预测方法引进负荷预测,解决了发展系数较大时不能用的禁限,最后对负荷预测差值进行适当阶次的对数处理,建立了局部残差处理模型。通过实例分析表明该方法较普通GM(1,1)模型具有更高的精度。

Load Forecasting Using Revised GM(1,1) Model with White Exponential Law Coincidence Property

The modeling mechanism and existing restriction of the traditional grey model are analyzed and a revised method is proposed.The original load data sequence is pre-processed by logarithm to reduce the stepwise ratio deviation.The white exponential law coincidence property is adopted into the load forecasting method to solve the problem that the model cannot be used when the developing coefficient is large.The load forecasting difference is processed by logarithm with appropriate order to establish local residual error processing model.Case studies show that this method has higher precision than general models.