Because of the problems of the various sources of uncertainties, wind power point forecasting models may lead to risks for power system operation and planning. The uncertainty sources include input uncertainty, model uncertainty, parameters uncertainty, and so forth. In general, in order to enhance the reliability and credibility of the wind power forecasting model outputs, prediction interval forecasting instead of point forecasting is used and provides a range of future values. However, for wind power prediction interval forecasting models, conventional methods for building prediction interval suffer from the assumption of data distribution, large computational complexity or difficult computation, resulting in generating inappropriate prediction interval. Besides, because of increased uncertainty from longer forecasting steps, constructing effectively the prediction interval of multistep-ahead forecasting is an important issue. Therefore, in this paper, we develop a center-of-concentrated-based neural network method for building the prediction interval of wind power forecasting systems in order to avoid the restrictive condition of data distribution, and the problem of difficult computation. Moreover, simulation results using different neural networks and heuristic optimizations are compared and analyzed to illustrate the effectiveness and feasibility of the proposed method.
ASJC Scopus subject areas