Short-term bus load forecasting based on XGBoost and Stacking model fusion

被引：0

作者：

Liu B. ^{[1
]}

Qin C. ^{[1
]}

Ju P. ^{[1
]}

Zhao J. ^{[2
]}

Chen Y. ^{[1
]}

Zhao J. ^{[2
]}

机构：

[1] College of Energy and Electrical Engineering, Hohai University, Nanjing

[2] State Grid Jiangsu Electric Power Company Research Institute, Nanjing

[3] State Grid Nanjing Power Supply Company, Nanjing

来源：

Dianli Zidonghua Shebei/Electric Power Automation Equipment | 2020年 / 40卷 / 03期

基金：

中国国家自然科学基金;

关键词：

Bus load; Meta-model; Particle swarm optimization algorithm; Stacking model fusion; XGBoost;

D O I：

10.16081/j.epae.202002024

中图分类号：

学科分类号：

摘要：

Bus load forecasting plays an important role in safe and stable dispatching of power grid, but bus load has strong stochastic fluctuation, and its attributes are various because of the difference of power supply areas, for which, a short-term bus load forecasting method based on XGBoost(eXtreme Gradient Boosting) and Stacking model fusion is proposed. Multiple bus load foresting meta-models are built based on XGBoost to form the meta-model layer of Stacking model fusion, an XGBoost model is connected to the meta-model layer for fusion, and the comprehensive forecasting system is formed. The particle swarm optimization algorithm is adopted to optimize the system parameters. The cases of 220 kV bus with different load attributes are analyzed, and the validity and applicability of the proposed method are verified. © 2020, Electric Power Automation Equipment Editorial Department. All right reserved.

引用

页码：147 / 153

页数：6

共 19 条

[1]

Wang Y., Wu M., Medium and long term load forecasting model based on partial least-square regression under new normal economy, Electric Power Automation Equipment, 38, 3, pp. 133-139, (2018)

[2]

Zu X., Tian M., Bai Y., Short-term load forecasting based on fuzzy clustering and functional wavelet-kernel regression, Electric Power Automation Equipment, 36, 10, pp. 134-140, (2016)

[3]

Li B., Men D., Yan Y., Et al., Bus load forecasting based on numerical weather prediction, Automation of Electric Power Systems, 39, 1, pp. 137-140, (2015)

[4]

Amjady N., Short-term bus load forecasting of power systems by a new hybrid method, IEEE Transactions on Power Systems, 22, 1, pp. 333-341, (2007)

[5]

Han Y., Li H., Bus load forecasting based on wavelet transform and SVM, Electric Power Automation Equipment, 32, 4, pp. 88-91, (2012)

[6]

Yin X., Xiao X., Sun X., Bus load forecasting model selection and variable weights combination forecasting based on forecasting effectiveness and Markov chain-cloud model, Electric Power Automation Equipment, 35, 3, pp. 114-119, (2015)

[7]

Panapakidis I.P., Clustering based day-ahead and hour-ahead bus load forecasting models, International Journal of Electrical Power & Energy Systems, 80, pp. 171-178, (2016)

[8]

Li G., Liu W., Yun H., Et al., A new data preprocessing method for bus load forecasting, Power System Technology, 34, 2, pp. 149-154, (2010)

[9]

Chen X.Y., Kang C.Q., Tong X., Et al., Improving the accuracy of bus load forecasting by a two-stage bad data identification method, IEEE Transactions on Power Systems, 29, 4, pp. 1634-1641, (2014)

[10]

Wang Y., Li D., Kang C., Application of low-rank matrix factorization in bad data identification and recovering for bus load, Power System Technology, 41, 6, pp. 1972-1979, (2017)

← 1 2 →