Research on the Rationality of Sample Size in Wind Power Forecast Evaluation Index

Abstract

On the basis of analyzing composition and distribution characteristics of wind power forecast, the paper put forward an analysis method of the minimum sample size by using the methods of normal distribution and hypothesis testing. The example shows, in order to reduce the probability of two types of errors at the same time, the capacity of the sample must be increased, the higher the confidence level needs the larger sample size. If the predicted errors is close to the assessment indicators, sample size need to be increased if not want to reduce the credibility.