基于数据挖掘与支持向量机的现货市场出清价预测方法  被引量:20

Forecasting Method of Spot Market Clearing Price Based on Data Mining and Support Vector Machine

在线阅读下载全文

作  者:陈杰尧 陶春华 马光文[1,2] 陈仕军[1,2] 赵永龙 王靖 CHEN Jieyao;TAO Chunhua;MA Guangwen;CHEN Shijun;ZHAO Yonglong;WANG Jing(College of Water Resource and Hydropower,Sichuan University,Chengdu 610065,Sichuan,China;State Key Laboratory of Hydraulics and Mountain River Engineering,Sichuan University,Chengdu 610065,Sichuan,China;Sichuan Dahui Big Data Service Co.,Ltd.,Chengdu 610041,Sichuan,China;Southwest Branch of State Grid Corporation of China,Chengdu 610041,Sichuan,China)

机构地区:[1]四川大学水利水电学院,四川成都610065 [2]四川大学水力学与山区河流开发保护国家重点实验室,四川成都6105065 [3]四川大汇大数据服务有限公司,四川成都610041 [4]国家电网公司西南分部,四川成都610041

出  处:《电网与清洁能源》2020年第10期14-19,27,共7页Power System and Clean Energy

基  金:国家重点研发计划(2018YFB0905204);国家电网有限公司科技项目(SGSCDK00XTJS1700047)。

摘  要:现货市场环境下,市场出清价对于电力市场的发、用两侧参与者和市场管理者都是极为重要的信息。因此,市场出清价的预测研究越发重要。首先分析大多数传统电价预测方法采用的连续序列与该文选取的同时段的电价序列表现出的变化特征差异,给出选取同时段电价序列作为输入的原因。然后基于数据挖掘相似性理论,通过欧氏距离和角度距离2个维度识别历史电价相似序列,得到模型所需训练集数据。以支持向量机(SVM)为预测工具,并利用遗传算法对SVM的关键参数进行寻优。最后将模拟预测结果与不考虑历史相似状态的SVM模型、BP神经网络模型进行对比,通过误差分析证明了所提模型具备更高的预测精度。The clearing price of the electricity spot market is the most useful information for both sides of the electricity market as well as market managers. Therefore,the research on forecasting the clearing price is particularly important. Different from the traditional electricity price prediction which uses a continuous sequence,this article uses a time-phased method to predict electricity prices in different time periods. First,based on the similarity theory of time series,historical similarity state sequences are identified through the two dimensions of Euclidean distance and angular distance to obtain the training set data required by the model. With support vector machine(SVM)as a prediction tool,the genetic algorithm is then used to optimize the key parameters of SVM, and finally the electricity price prediction results in different periods can be obtained. The article uses the Nordic electricity market data to make simulation predictions,and compares the results with the SVM models and BP neural network models that do not consider historical similar states,and the error analysis shows that the proposed model has higher prediction accuracy.

关 键 词:市场出清价预测 支持向量机 遗传算法 相似性原理 同时段电价预测 

分 类 号:TM715[电气工程—电力系统及自动化]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象