A discrete point estimate method for probabilistic load flow based on the measured data of wind power

Abstract

Probabilistic load flow (PLF) calculation is the first step to evaluate the impact of the integrated wind power to the power system. The wind power is featured with stochastic and variable properties and it's hard to fit its distribution characteristics to any common probability density function (PDF). However, the traditional methods including Monte Carlo for PLF are based on the input variable's PDF. In the paper, the point estimate method and Gram-Charlier expansion method are combined. Based only on the sample data of the wind power, the expectation, variance and cumulative distribution of the output random variables can be estimated with the method by 2n+1 times of load flow calculation where n is the number of input stochastic variables, exempting the need for distribution of the input variables. The simulation results in the IEEE 16-generator system show that the method provides high precision with less computation burden. The method can also be applied to other problems with uncertainty factors whose distribution is unknown in the power system.