Dynamic equivalent modeling of wind farm considering the uncertainty of wind power prediction and a case study

Abstract

Currently, the clustering of dynamic multi-machine equivalent modeling of wind farm (WF) is based on the deterministic data. To serve the early warning calculation of power system dispatch better, a dynamic multi-machine equivalent modeling method of WF considering the uncertainty of wind power prediction is first proposed in this paper. First, the probability density functions of wind speed and wind power prediction errors are described and the multi-scenario sampling method is adopted to reflect the uncertainty. Second, the probabilistic clustering method is used to obtain the probability of each wind turbine generator (WTG) belonging to each cluster. The cluster with the biggest probability is selected as the clustering result of the WTGs in the spirit of the Bayesian classification. Third, the WTGs and collector system are simplified into a single-machine equivalent model in each cluster. Finally, the WF output response curve is compared among the detailed WF models, the conventional and the proposed equivalent models in the accurate and inaccurate scenarios of wind power prediction, respectively. The results show that the proposed equivalent method considers the uncertainty in reality and has a better accuracy in the inaccurate scenario of wind power prediction.