In order to improve the computing efficiency of probabilistic analysis in an unbalanced three-phase power system

a probability analysis and radial power flow analysis method is proposed by considering the dispersed generator load demand and the injected power uncertainty. Firstly

the three-point estimate algorithm is improved with Nataf transform

so that the algorithm can deal with incomplete multi-variable information

and the correlated random variables are used to improve the accuracy of the estimation of output variables. Secondly

the dispersion power network model is studied

with which the Nataf transform three point estimation algorithm is used for realizing the analysis of power flow calculation and short circuit. Finally

with Monte Carlo simulation method

the proposed algorithm is verified on an improved IEEE 57-bus test system

the results showing that the cumulative distribution function of voltage amplitude for the bus 51 and the random variable of the output are both more accurate than those with the three point estimation algorithm. These advantages also exist in the short circuit analysis. This work is supported by National Natural Science Foundation of China (No. 51377111).