|
摘要: |
提出了一种基于主成分分析(PCA)的最小二乘支持向量机(LS-SVM)短期负荷预测模型,模型中引入多元统计分析中的主戍分分析理论来解决输入变量的选择问题。该模型首先对样本的高维变量数据矩阵进行标准化处理,建立相关矩阵,计算特征值和特征向量,然后求取累计方差贡献率,并据此求取主成分作为最小二乘支持向量机的输入进行训练预测。主成分以较少的维数包含了原高维变量所携带的大部分信息,全面地考虑了影响负荷预测的各种因素,又避免了过多的输入导致的精度低、训练慢的不足。实例表明,所提方法可有效地消除众多影响因素间的相关性,减少输入变量个数,提高预测效率和精度。 |
关键词: 短期负荷预测 影响因素 主成分分析 最小二乘支持向量机 |
DOI: |
分类号:TM715 |
基金项目:国家科技支撑计划项目
? |
|
Short-term load forecasting model based on LS-SVM with PCA |
LIU Baoying YANG Rengang
|
Abstract: |
A short-term load forecasting model based on LS- SVM(Least Squares Support Vector Machines)is presented by using PCA(Principle Components Analysis,a method of multivariate statistic analysis)to select input variables.The original high-dimensional data matrix is normalized to establish the correlation matrix for the calculation of eigenvalues,eigenvectors and accumulated contribution of variance,according to which the principal components are determined.The low -dimensional principal components with most information included in the original high-dimensional data set are used as the inputs of LS-SVM for training and prediction.The forecasting model considers all-sided influencing factors and avoids the low precision and slow training induced by over-input. The example shows that it eliminates the relevance among factors,reduces the input variables and improves the accuracy and efficiency. |
Key words: short-term load forecasting,influencing factors,principal component analysis,least squares support vector machines |