针对中长期电力负荷预测样本量小、多因素影响的特点,利用灰色关联度筛选影响因素,建立基于BP

神经网络算法的负荷预测模型,通过多因素变量及历史负荷变量序列进行滚动预测,得到的预测值明显优于

单一预测方法,并通过马尔可夫过程对预测残差进行修正,使预测精度得到较大提高,研究实证表明,这种预

测方法具有进行推广应用的价值

-For long-term load forecasting small sample size, the characteristics of many factors, the use of gray correlation screening factors is established based on BP

Neural network algorithm for load forecasting model, the multivariable and historical load variable sequences rolling forecasts, significantly better than the predicted values ​ ​ obtained

Single forecasting methods, and through the Markov process to amend the prediction residuals, the prediction accuracy is greatly improved, empirical studies show that pre-

Method for measuring the value of popularization and application