基于相空间分析的电力负荷序列预测算法

Power Load Forecasting Algorithm Based on Phase Space Analysis

对电力负荷时间序列进行准确预测,避免电力负荷过载,传统方法采用线性时间序列预测算法,没有考虑电力负荷时间序列的自相似性和高斯性,导致预测精度不高。提出一种基于非线性时间序列分析和相空间重构的电力负荷序列预测算法。构建了电力负荷数据的非线性时间序列分析模型,采用相空间重构方法把电力负荷数据嵌入到高相空间中,在相空间中提取电力负荷时间序列的非线性特征,实现对电力负荷时间序列的准确预测,仿真结果表明,采用该方法进行电力负荷预测的精度较高,通过相空间分析能有效反应电力负荷数据的内部结构特征,提高预测性能和电力负荷数据的特征分析能力。

The accurate prediction of power load time series can avoid the power overload.The traditional method uses linear time series forecasting algorithm,without considering the self similarity and Gauss of power load time series,and the prediction accuracy is not high.This paper presents a power load forecasting algorithm based on nonlinear time series analysis and phase space reconstruction.The nonlinear time series analysis model of power load data is constructed,and the nonlinear characteristics of power load data are extracted by using phase space reconstruction method.The simulation results show that the accuracy of power load forecasting is high.The analysis of the phase space can effectively reflect the features of the internal structure of electric power load data,and improve the prediction performance and the characteristic analysis ability of the power load data.