摘要： 电价的分类与预测是电力市场电价理论研究中的重要内容。该文提出了混合贝叶斯支持向量机方法(BE-SVM)，通过贝叶斯统计方法对电价进行分类，挖掘有效的数据信息，并结合支持向量机(SVM)技术预测现货电价数据，贝叶斯前验分布和后验分布用来估计SVM中的参数。通过比较模型BE-SVM、SVM 和神经网络(ANN)的预测结果，表明该文提出的BE-SVM方法提高了电价的预测精度，是一种有效的方法。

关键词: 贝叶斯分类, 支持向量机, 市场电价, 参数估计

Abstract: Electricity market price classification and forecast are important elements in the theoretical study. This paper proposes a hybrid numeric Bayesian with support vector machine(BE-SVM)method that integrates a Bayesian statistical method for electricity price classification with experience distribution approach in data information mining, and SVM technique for electricity forecasting. The Bayesian prior distribution and posterior distribution are used to evaluate the parameters in the SVM. Experimental results show that the proposed BE-SVM method has a higher forecast accuracy compared with three models BE-SVM, SVM and ANN.

Key words: Bayesian classification, support vector machine(SVM), electricity market price, parameter estimation

中图分类号: 

• TP311