A multiple-variable simulation method based on the Johnson system is proposed and applied in the annual reliability evaluation of a generation system with wind farms. By using the historical data of the wind speed and load duration curve

the relationship between wind speed and standard normal variable(SNV)and that between load and SNV are developed in the method

i. e. the characteristics of probability distribution of wind speed and load are simulated. Accordingly

correlated wind speeds are simulated by simulating correlated SNVs

thus avoiding possible low accuracy of simulation due to the assumption about the wind speed probability distribution model. The accuracy of the method in simulating the wind speed and load duration curve is verified

as is the adaptability in reliability evaluation of generation systems with wind farms. Finally

the influence of wind speed correlation on generation system reliability is analyzed.This work is supported by Key Program of National Natural Science Foundation of China(No. 51337005)and State Grid Corporation of China(No. XT71-14-002).