A Probabilistic Method To Forecast The Regional And Mid-long Term Wind Power Generating Capacity

In recent years, the development of wind power in China is sustained and rapid. However, the characteristics of the distribution of wind resources is centralized, large-scale and far away from the load center. As a result, a lot of wind power is difficult to local accommodates and appears a serious phenomenon of curtailed wind power. Taking the characteristics of complementarity between regional resources into account, the wind power is transported to load centers through large-scale, long-term power trading, so as to achieve the purpose of wind power accommodation. In the implementation of cross-regional, long-term electricity trading to address the large-scale wind power accommodation pathways, regional accurate prediction of long-term wind power generating capacity is a key part.Compared to the short-term wind power prediction, there are some difficulties of regional and mid-long term wind power generating capacity prediction:Due to the short history of wind power generation, historical annual generating capacity of wind power and the historical wind measurement data of wind farms is very few; Due to the rapid development of wind power in China, the newly installed capacity of wind farms is large and unregularly.This paper starts with the annual probabilistic distribution prediction of wind power, based on the mapping of wind energy and wind power, obtains the results of probabilistic prediction of regional mid-long term wind power generating capacity of temporarily excluding the influence of new capacity; This paper predicts the annual average utilization hours of wind power, and multiplies the result above with planning capacity to obtain the newly installed generating capacity of the forecasted year, then adding the probability of the predicted results with newly installed generation capacity to get the final prediction.The above ideas still encounter the difficulty of direct forecasting of regional wind power generating capacity. The adopted solution is to select reference wind farm in the regional wind farms, and predict the generating capacity of the reference wind farm, then get the regional probability results of wind power generating capacity by the extending model.