Research on Power Network Load Forecasting Problem Based on Machine Learning. Issue 1 (March 2020)

• Record Type: Journal Article
• Authors: Wang, Zhankui; Wu, Junying; Xin, Rui; Bai, Tao; Zhao, Jianbin; Wei, Minglei; Zhuang, Lei
• Is Part Of: IOP conference series. Volume 453:Issue 1(2020)
• Publication Date: 2020-03
• Subjects: Electric load -- support vector machine -- particle swarm optimization.; Earth sciences -- Periodicals; Environmental sciences -- Congresses; Environmental sciences -- Periodicals; 550.5
• DOI: 10.1088/1755-1315/453/1/012055
• Languages: English
• ISSNs: 1755-1307

Abstract

Abstract: Based on the existing electric power grid load forecasting method and improving the precision of electric power load forecasting, an power load forecasting model based on Support Vector Machine (SVM) and improve the parameters by Particle Swarm Optimization (PSO). Firstly, analyzed the theoretical basis of support vector machine, and the preliminary prediction model of support vector machine is established. Then, the PSO algorithm is used to iteratively choose the optimal parameters of the support vector machine parameters. Finally, the optimal load prediction model is established by the optimal parameters. The pre-processed real power data is input into the model for learning prediction, and the model prediction effect is verified by Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE) and prediction chart. The experimental results show that the PSO-SVM prediction model can precise forecast the power load and raised the accuracy of power load forecasting. It shows that the PSO algorithm is very effective to adjust the parameters of SVM model.