一种基于粗糙集理论的SVM短期负荷预测方法

在分析粗糙集理论方法与支持向量机方法的优势和互补性后,探讨了粗糙集与支持向量机的结合方法,提出了一种基于粗糙集数据预处理的支持向量机预测系统。该系统利用粗糙集理论在处理大数据量、消除冗余信息等方面的优势,减少支持向量机的训练数据,克服支持向量机方法因为数据量太大,处理速度慢等缺点。将该系统应用于短期负荷预测中,与BP神经网络法和标准的支持向量机方法相比,得到了较高的预测精度,从而说明了基于粗糙集理论方法作为信息预处理的支持向量机学习系统的优越性。

Approach to forecast short-term load of SVM based on rough sets

After analyzing advantage and complementarily of the methods of rough sets (RS) theory and support vector machines (SVM) making an inquiry into combining method of RS with SVM, a kind of SVM forcasting system based on rough sets data preprocess is proposed. Utilizing the advantages of RS theory in processing large data and eliminating redundant information the system has decreased SVM training data and overcome the disadvantages of very large data and slow processig speed caused by SVM approach. In using forecasting short-term load this approach has achieved greater forcasting accuracy comparing with the method of BP neural network and standard SVM. It is denoted that the SVM learning system has advantage as the information preprocesing based on RS approach.