With the development of renewable energy，the probability distribution of combined output of renewable energy generation becomes an outstanding issue in the power system.This paper summarizes the distribution characteristics of several common renewable energy，and proposes the cumulative distribution function and probability density function of the combined output.In addition，two methods are proposed to get the confidence intervals:dichotomy search numerical integration method and Latin hypercube sampling Monte Carlo method.Numerical tests are conducted with a microgrid system.Testing results indicate that the proposed formula is accurate.Additionally，dichotomy search numerical integration method has higher accuracy but has a time consumption of228.25 s，while sampling methods need less than1s.Compared to random sampling，Latin hypercube sampling Monte Carlo method has higher sampling efficiency，higher accuracy and lower time consumption.