Research of Probabilistic Modeling Method Aimed at Uncertain Factors

Abstract

the rapid development of distributed generation technology has brought a large number of uncertain factors to the power grid. The research of probability distribution model of these uncertain factors has great significance to the planning and operation of the power grid. Taking the uncertainty problem in wind farm as an example, a non-parametric kernel density estimation method for wind power probability density model based on fuzzy ordinal optimization is proposed in this paper. Based on the historical data sample of wind power operation, a non-parametric kernel density estimation model of wind power probability density is established in the method, and then a multi-objective optimization model of bandwidth selection is also constructed. Finally, the fuzzy ordinal optimization is used to solve the bandwidth optimization model. The result of practical examples show that the method is completely driven by the sample data and has no need to perform a priori subjective assumption on the probability density model as in the traditional way. Therefore, it has higher modeling accuracy and stronger applicable ability.