新能源配电网低碳运行长期负荷预测研究

Research on long-term load forecasting for low-carbonoperation of newenergy distribution networks

为解决新能源配电网持续低碳运行时的长期负荷预测难度问题。提出利用多时间尺度分析法获取电力负荷数据的小波尺度,选取多个尺度进行分析获取的初步长期负荷数据,采用指数平滑法对获取的初步长期负荷数据进行平滑处理。通过对平滑处理后长期负荷数据的累加累减以及还原,构建一阶微分方程,利用灰色关联方法实现配电网低碳运行长期负荷预测。结果表明,所提方法的预测结果接近电力负荷实际值,预测误差始终低于0.3 MW,且在配电网不同时期的预测误差范围均较小。

In order to solve the problem of long-term load prediction in the continuous low-carbon operation of new energy distribution network.The wavelet scale of power load data obtained by using multi-time scale analysis method was proposed,the preliminary long-term load data obtained by analyzing multiple scales was selected,and the preliminary long-term load data obtained by exponential smoothing method was smoothed.Through the accumulation,subtraction and reduction of the smoothed long-term load data,the first-order differential equation was constructed,and the grey correlation method was used to realize the long-term load prediction of low-carbon operation of the distribution network.The results showed that the prediction results of the proposed method were close to the actual value of power load,and the prediction error was always less than 0.3 MW,and the prediction error range was small in different periods of the distribution network.