基于改进时间序列法的配电网短期负荷预测模型

Short-term Load Forecast Based on Time-series in Distribution System

探讨了配电网数据预处理的问题 ,采用了首端功率系数法将非实测节点的数据补充完整 ,并建立了改进的时间序列法 ,避免了大量复杂的计算 ,包括矩阵的转置所带来的不便 ,实现了配电网短期负荷预测模型的在线预测。

This paper thoroughly researches the method on data pretreatment for short-term load forecasting in distribution system,introduces sending-side power coefficient to attain the data of non-measured nodes An improved time-series method is presented,which avoid the complex calculation including matrix inversion for parameter estimations,the algorithm is computationally efficient Because of this high efficiency,our forecasting model can be updated daily or even hourly