: It is important to model the future scenarios of wind farm power output in enhancing capability of wind power accommodation and decreasing wind power curtailment. A scenario generation strategy considering the correlation between multiple wind farms is proposed. A convolutional neural network combined with quantile regression technique is introduced to achieve detailed quantiles of corresponding predicted wind power output, which can be regarded as the cumulative distribution function (CDF) by approximation. Marginal conditional probability density function (PDF) for each wind farm can be constructed from the CDF. A copula function-based method is used to form the joint PDF of multiple wind farm power outputs from the marginal PDF constructed before. By inversing the joint PDF, the required scenario set can be formed. In case studies, the proposed strategy is tested with two wind farms data, and the simulation results verify the effectiveness of the proposed strategy.
[1]
James W. Taylor.
A Quantile Regression Neural Network Approach to Estimating the Conditional Density of Multiperiod Returns
,
2000
.
[2]
Lu Zongxiang.
Unit Commitment Based on Wind Power Forecast
,
2011
.
[3]
M. O'Malley,et al.
Stochastic Optimization Model to Study the Operational Impacts of High Wind Penetrations in Ireland
,
2011,
IEEE Transactions on Power Systems.
[4]
R. Koenker,et al.
Regression Quantiles
,
2007
.
[5]
Yann LeCun,et al.
Convolutional networks and applications in vision
,
2010,
Proceedings of 2010 IEEE International Symposium on Circuits and Systems.
[6]
H. White.
Nonparametric Estimation of Conditional Quantiles Using Neural Networks
,
1990
.