摘要
电力系统的输电量预测是电网调度人员的参考指标之一,输电量预测的准确性与时效性在一定程度上影响着用户的用电质量。电力系统运在行过程中,每天都会产生大量的运行数据,为了利用这些大数据以及更好地实现输电线路输电量预测精度和快速性,提出了基于极限学习机的输电线路输电量预测模型。通过实验表明,该模型有良好的预测精度,对不同输电线路的输电量数据有着不错的泛化能力。通过与小波神经网络预测结果的对比,所建模型的训练和测试时间比小波神经网络快了约67s,尤其模型的训练时间极短,该模型预测的MAPE值要比小波神经网络低9%左右。
Power transmission volume forecast is one of the reference indexes for grid dispatchers.The accuracy and timeliness of the forecast affect the power quality of the users.During the operation of power system,a large amount of operation data will be generated every day.In order to make use of these big data and better realize the accuracy and rapidity of transmission line transmission volume prediction,this paper proposes a transmission line transmission volume prediction model based on extreme learning machine.Experiments show that the model has good prediction accuracy and good generalization ability for transmission volume data of different transmission lines.Compared with the prediction results of the wavelet neural network,the training and testing time of the model is nearly 67 sfaster than that of the wavelet neural network.In particular,the training time of the model is extremely short,and the predicted MAPE value of the model is about 9%lower than that of the wavelet neural network.
作者
郑熠旻
ZHENG Yi-min(Putian Power Supply Company,State Grid Fujian Electric Power Company,Putian351100,China)
出处
《电力学报》
2019年第4期354-362,共9页
Journal of Electric Power
关键词
电力系统
输电量预测
极限学习机
小波神经网络
power system
transmission volume forecast
extreme learning machine
wavelet neural network
