Yanbian Power Supply Company

Boji, China

Yanbian Power Supply Company

Boji, China
SEARCH FILTERS
Time filter
Source Type

Wang X.,Shanghai JiaoTong University | Huang K.,Chongqing Electric Power Company | Zheng Y.,Shanghai JiaoTong University | Li L.,Shanghai JiaoTong University | And 3 more authors.
Dianwang Jishu/Power System Technology | Year: 2017

With increasing larger capacity of PV generation integrated into power grid, its output randomness is bound to pose certain effect on safe and stable operation of power grid. A new short-term combined forecast model with variable weights was proposed in this paper. Firstly, to simplify model input dimensions, multiple linear factors influencing PV output were compressed and extracted with principal component analysis (PCA) method. Then the first principal component extracted from PCA combined with grey correlation degree was used to filter similar historical days. Next, the chosen days were respectively brought into two models, least square support vector machine (LS-SVM) and modified BP network (MBP), and two predictions were repeated: the first was forecast for similar day and then firefly algorithm for generalized regression neural network (FFA-GRNN) was applied to train weight coefficients; the second was ultimate forecast for test sets. Simulation results show validity of the proposed model. © 2017, Power System Technology Press. All right reserved.

Loading Yanbian Power Supply Company collaborators
Loading Yanbian Power Supply Company collaborators