Short-term load forecasting based on fuzzy clustering and functional wavelet-kernel regression

被引：0
|
作者：
Zu X. [1 ]
Tian M. [1 ]
Bai Y. [1 ]
机构：
[1] Department of Control & Computer Engineering, North China Electric Power University, Beijing
来源：
| 2016年 / Electric Power Automation Equipment Press卷 / 36期
基金：
中国国家自然科学基金;
关键词：
Fuzzy clustering; Nonlinear time series; Pattern similarity method; Short-term load forecasting; Smart grid; Wavelet-kernel nonparametric regression;
D O I：
10.16081/j.issn.1006-6047.2016.10.020
中图分类号：
学科分类号：
摘要：
STLF(Short-Term Load Forecasting) is an important issue of smart grid. The historical daily load curve is expressed as a set of hourly load segments and a hybrid forecasting algorithm based on the pattern similarity method is applied, which combines the fuzzy clustering with the FWKNR(Functional Wavelet-Kernel Nonparametric Regression). FNWKR is applied to express the load curve of the predicted day as a set of weighted average of the corresponding hourly load segments of historical days, which assigns higher weight to the segment with higher similarity and uses N-WE(Nadaraya-Watson Estimator) to calculate the weight based on the shape-similarity measurement of discrete wavelet transform. The daily load is predicted by the quick forecasting of load segments. Fuzzy clustering is used to pre-classify the historical loads to typical load patterns for a particular customer and recognize the effective reduced training sample set with more similar behaviour pattern to the predicted day for the model forecasting. Based on the practical load data of a region, the experimental analysis verifies the superiority of the proposed algorithm. © 2016, Electric Power Automation Equipment Press. All right reserved.
引用
收藏
页码：134 / 140and165
相关论文
共 26 条
  • [1] Wang D., Sun Z., Big data analysis and parallel load forecasting of electric power user side, Proceedings of the CSEE, 35, 3, pp. 527-537, (2015)
  • [2] Zhong Q., Sun W., Yu N., Et al., Load and power forecasting in active distribution network planning, Proceedings of the CSEE, 34, 19, pp. 3050-3056, (2014)
  • [3] Da Silva P.G., Ilic D., Karnouskos S., The impact of smart grid prosumer grouping on forecasting accuracy and its benefits for local electricity market trading, IEEE Transactions on Smart Grid, 5, 1, pp. 402-410, (2014)
  • [4] Quilumba F.L., Lee W.J., Huang H., Et al., Using smart meter data to improve the accuracy of intraday load forecasting considering customer behavior similarities, IEEE Transactions on Smart Grid, 6, 2, pp. 911-918, (2015)
  • [5] Kwac J., Flora J., Rajagopal R., Household energy consumption segmentation using hourly data, IEEE Transactions on Smart Grid, 5, 5, pp. 420-430, (2014)
  • [6] Chaouch M., Clustering-based improvement of nonparametric functional time series forecasting: application to intra-day house-hold-level load curves, IEEE Transactions on Smart Grid, 5, 1, pp. 411-419, (2014)
  • [7] Feinberg E.A., Genethliou D., Applied Mathematics for Restructured Electric Power Systems: Optimization, Control, and Computational Intelligence: Load Forecasting, pp. 269-285, (2005)
  • [8] Lei S., Sun C., Zhou Q., Et al., Method of multivariate time series of short-term load forecasting, Transactions of China Electrotechnical Society, 20, 4, pp. 62-67, (2005)
  • [9] Wang X., Zhang S., An improved method for short-term electric load forecasting using time series techniques, Journal of Shanghai University(Natural Science), 8, 2, pp. 133-136, (2002)
  • [10] Chen H., A new load forecasting method based on autoregressive conditional heteroscedasticity model, Automation of Electric Power Systems, 31, 15, pp. 51-54, (2007)
123