Abstract: In the development trend of large-scale wind power connected into power system, power system will be affected by more complicated random factors, which bring a huge challenge to the security and stability of operation. By considering stochastic disturbance of wind farms under different scales and different outlets, stochastic outage rate of generator and stochastic volatility of load, a new evaluation method of probabilistic vulnerability for transmission network is proposed based on Cumulant and Gram-Charlier series (CGC), by combing with critical voltage and ultimate transmission power of the system. In this way, the weak nodes and branches of power system connected with wind farms are screened out. The proposed method is verified by the simulation of IEEE-30 bus system under different scale wind farms and different access points, and the results are compared with that obtained by traditional evaluation methods based on branch potential energy function method. The results show that the proposed method can provide reliable and effective evaluation results, and it is helpful to further explore the vulnerability of power system with different scale wind farm integration.