摘要
随着我国电力体制改革的不断深入,售电公司作为电力市场的主要参与者,其主要获利方式是从电力市场中购买电量并销售给用户。因此准确预测现货市场电价变化趋势,是售电公司降低购售电风险的重要保障。为此,根据现货市场中电价的特性,提出基于集成经验模态分解(ensemble empirical mode decomposition,EEMD)、支持向量机(support vector machine,SVM)和自回归移动平均模型(autoregressive moving average,ARMA)的组合预测模型。首先利用EEMD将历史数据分解成一系列相对比较平稳的分量序列;其次,利用遗传算法(genetic algorithm,GA)优化的SVM预测高频分量,利用自回归移动平均模型预测低频分量;最后将各子序列的预测结果求和作为最终预测结果。用美国售电公司真实数据进行预测,并与其他模型进行比较。算例结果表明所提模型的预测精度更高。
With the continuous deepening of China’s electric power system reform, as a major participant in the electric power market, the main way for electric power company to profit is to buy electricity from the electric power market and sell it to users. Therefore, accurately predicting the change trend of spot market electricity price is an important guarantee for electric power companies to reduce the risk of electricity buying and selling. For this purpose,based on the characteristics of electricity prices in the spot market,composite prediction models is proposed based on integrated empirical mode decomposition(EEMD), support vector machine(SVM)and autoregressive moving average(ARMA). Firstly, EEMD is used to decompose historical data into a series of relatively stable component sequences. Secondly, the SVM optimized by genetic algorithm(GA) is used to predict the high-frequency components, and the autoregressive moving average model is used to predict the low-frequency components. Finally, the final prediction result is the sum of the predicted results of each subsequence. The prediction is made using real data from American electric power companies and compared with other models. The numerical results show that the forecasting accuracy of the proposed model is higher.
作者
张金良
王明雪
ZHANG Jinliang;WANG Mingxue(Economics and management College,North China Electric Power University,Beijing 102206,China)
出处
《电力需求侧管理》
2020年第3期63-68,共6页
Power Demand Side Management
基金
国家自然科学基金(71774054)
中央高校基本科研业务专项资金项目(2019MS055)。
关键词
电价预测
集成经验模态分解
支持向量机
自回归移动平均模型
electricity price forecasting
integrated empirical modal decomposition
support vector machine
autoregressive moving average model
作者简介
张金良(1981),男,江苏常州人,副教授,主要研究方向为技术经济评价;王明雪(1995),女,河北沧州人,硕士研究生,主要研究方向为电价预测。
