论文部分内容阅读
根据电价时间序列的混沌特性,结合混沌理论和支持向量机方法提出了一种新的电价预测模型。该模型基于混沌理论对电价时间序列进行相空间重构,并根据相空间演变规律确定模型的输入输出结构,然后采用支持向量机拟合相点演化的非线性关系。为增强模型的泛化推理能力,训练样本按照预测相点最近邻点原理选择。对美国PJM电力市场边际电价历史数据的仿真研究表明,文中提出的预测模型能有效、稳定地提高电价预测精度。
According to chaotic characteristics of time series of electricity prices, a new electricity price forecasting model is proposed based on chaos theory and support vector machine. Based on the chaos theory, this model reconstructs the phase space of electricity price time series, and determines the input and output structure of the model according to the law of phase space evolution. Then the nonlinear relationship of phase evolution is fitted by using support vector machine. In order to enhance the generalization reasoning ability of the model, the training samples are selected according to the nearest neighbor point principle of the predicted phase point. The simulation research on the historical data of the marginal electricity price in PJM power market in the United States shows that the forecast model proposed in this paper can effectively and steadily improve the forecasting accuracy of electricity prices.