Interval prediction for long-term wind power of wind farm clusters based on cloud theory
DOI：10.7667/PSPC180214
Key Words:interval prediction  GARCH-t model  D vine Pair Copula model  cloud theory
Author NameAffiliation
CHEN Haojie Shanghai University of Electric Power, Shanghai 200093, China 
CHENG Haozhong Key Laboratory of Power Transmission and Conversion Shanghai Jiao Tong University, Ministry of Education, Shanghai 200240, China 
XU Guodong Key Laboratory of Power Transmission and Conversion Shanghai Jiao Tong University, Ministry of Education, Shanghai 200240, China 
MA Zeliang East China Branch of State Grid Corporation of China, Shanghai 200002, China 
FU Yesheng East China Branch of State Grid Corporation of China, Shanghai 200002, China 
Hits: 3615
Download times: 1371
Abstract:The uncertainty of wind power poses great challenges to the cluster development and grid connection of wind power. The point prediction of wind power is difficult to meet the actual need of flexible planning of the power network. Regarding to the problem of interval prediction for long-term wind power of wind farms clusters, the D-vine Pair Copula-GARCH-t model based on cloud theory is proposed to predict the output range of wind farms. The GARCH-t model can accurately reflect the leptokurtic characteristics of prediction errors and thus improve the precision of wind power prediction. Meanwhile, the D-vine Pair Copula model effectively describes the correlation between the wind power of wind farm clusters. Based on such cloud model with the digital characteristics of expectation, entropy and hyper entropy, the interval wind power can not only reflect the randomness and fuzziness, but also describe reasonably the relationship between them. The interval prediction for wind power could provide a reference for planner to do the long-term planning of wind farms. This work is supported by National Natural Science Foundation of China (No. 51337005).
View Full Text  View/Add Comment  Download reader