基于变分模态分解的中期电力负荷混合预测模型

A hybrid prediction model of medium-term power load based on variational mode decomposition

电力负荷预测直接影响电网规划和运行,但是受到各类因素的影响。为提高预测精度,针对电力负数据时序性和非线性特征,提出一种基于变分模态分解的中期电力负荷混合预测模型(hybrid prediction model of medium-term power load based on variational mode decomposition,HPMMPL-VMD)。在HPMMPL-VMD算法中,首先使用VMD将原始电力负荷序列分解成若干个相对平稳的模态分量,并利用长短时记忆神经网络对各个模态分量进行建模;然后将各个预测分量进行叠加得到电力负荷预测值;最后,使用最小二乘支持向量回归对误差序列进行预测,并将电力负荷预测值与误差预测值相加得到最后预测结果。为验证HPMMPL-VMD的性能,选取其他预测方法与其进行比较,实验结果表明本文所提模型具有较高的预测精度。

The power grid planning and operation are directly affected by the power load forecasting.However,it is a difficult task to accurately predict power load due to the influence of various factors.In order to improve the prediction accuracy,a hybrid prediction model of medium-term power load based on variational mode decomposition(HPMMPL�VMD)is proposed for the characteristics of time-series and nonlinear nature of power load data.In the HPMMPL-VMD algorithm,the VMD is firstly used to decompose the original power load sequence into several relatively stationary mode components.Then,the long and short-term memory neural network is used to model each mode component.After that,the predicted components are superimposed to obtain the power load predicted value.Finally,the least squares support vector regression is used to predict the error sequences.The final prediction result is obtained by adding the power load prediction value and the error prediction value.In order to verify the performance of HPMMPL-VMD,other prediction methods are selected for comparison.The experimental results show that the proposed method has high prediction accuracy.