基于方差-协方差组合预测的中长期电力负荷预测研究

Research on Medium and Long-term Electric Load Forecasting Based on Variance-covariance Combined Model

关于中长期电力负荷预测有多种预测方法,但各方法被许多因素所局限,本文选用线性回归模型、二次多项式模型和灰色预测模型分别对电力负荷进行预测,再综合起来建立方差-协方差组合预测模型。通过比较后得出结论,本方法大大降低了误差,提高了预测精度,可以科学地预测中长期的电力负荷。

There are a variety of methods to predict long-term electric load. However, the limitations of the methods are a number of factors. The linear regression model, quadratic polynomial model and gray prediction model are proposed in the paper for power load forecasting, respectively. These thee methods are taken together again to establish the variance-covariance combination forecasting model. By comparison concluded, this method greatly reduces errors, improves the prediction accuracy and can scientifically predict long-term power load.