基于混沌时间序列的地区电网负荷预测  

Load Forecasting of a Regional Power Grid Based on the Chaotic Time Series

作　　者：董子晗 DONG Zihan(Construction Branch of State Grid Shaanxi Electric Power Company,Xi’an 710000,Shaanxi,China)

机构地区：[1]国网陕西省电力公司建设分公司

出　　处：《电网与清洁能源》2019年第5期38-41,共4页Power System and Clean Energy

摘　　要：通过采用改进的C-C方法计算得出电网日负荷的最优时延和最佳嵌入维数,从而进行相空间重构,并通过计算最大Lyapunov指数,指出该时间序列的混沌特性。然后采用基于奇异值分解的Volterra方法对某地区电网负荷进行预测。预测结果显示,该方法的预测效果较好,具有较高的预测精度,并能够反映电网未来负荷的变化趋势。The optimal latency and the best embedding dimension of the daily load of the power grid are calculated by the improved C-C method,both of which then are used to reconstruct the the phase-space.The largest Lyapunov index is calculated to describe the chaotic characteristics.Then the Volterra method based on singular value decomposition is used to predict the load of a regional power grid.The forecasting results show that the method works better in forecasting with high accuracy,and can reflect the trend of future load changes of the power grid.

关 键 词：混沌 相空间重构 VOLTERRA滤波器 负荷预测 

分 类 号：TM715.1[电气工程—电力系统及自动化]