相空间重构和支持向量机结合的电力负荷预测模型研究.doc

相空间重构和支持向量机结合的电力负荷预测模型研究* 李昕，，CAO方法分别求得时间延迟和嵌入维数，并由此得到系统最大李雅普诺夫指数，证明其具有混沌特性。然后根据时间延迟和嵌入维数对样本数据相空间重构，在此基础上利用支持向量回归算法（PSR-SVR）BP神经网络模型的预测结果进行对比，结果表明，这是一种误差小，精度高的电网负荷预测方法器。 关键词: 支持向量回归; 混沌; 相空间重构; 电力负荷预测 中图分类号: TM714 文献标识码: A 文章编号：1001-1390（2014）00-0000-00Study on Power Load Forecasting Model Based on Phase Space Reconstruction and SVM LI Xin, YAN Hong-wei, MA Hong-yi (School of Mechanics and Power Engineering, North University of China, Taiyuan 030051, China) : The impacts of centralized operation of wind turbines on grid’s security and stability operation ask for high requirements for power load forecasting precision in order to realized reasonable planning efficient operation of various power supply units. The time series of grid has chaotic characteristics and it is difficult to describe its characteristics and inherent laws. The chaotic phase space reconstruction theory is adopted to study the power load time series sample data. Time delay and embedding dimension are obtained through the mutual information method and the CAO. Lyapunov exponent of this system is obtained so as to prove that the grid system has chaotic characteristics. Then the phase space is reconstructed according to the time delay and embedding dimension. On the basis of phase space reconstruction, support vector regression algorithm is adoped to predict the power load. The grid search method is used for parameter optimization. Finally, the predicted results with the time series prediction model and BP neural network model are compared. The results show that is the proposed method is a high precison load forecasting method with small error. Key words： chaos theory, phase space reconstruction, support vector regression, power load forecasting 灰色预测模型…,N}，选取适当的嵌入维数m和延迟时间τ，就可以用相空间重构方法得到相空间中的数据集。 (1)式中 X(t) = [x(t), x(t+τ),…,x (t+(m-1) τ)]，Y(t)=x(t+1+(m-1) τ)，t=1,2,…,M ， X(t)为m维相空间中的相点，M为相点个数，且M=N-(m-1) τ。 写为矩阵形式即为 (2) (3)