This paper proposes a power network vulnerability analysis method based on topological approach considering of uncertainties from high-penetrated wind generations. In order to assess the influence of the impact of wind generation owing to its variable wind speed etc., the Quasi Monte Carlo based probabilistic load flow is adopted and performed. On the other hand, an extended stochastic topological vulnerability method involving Complex Network theory with probabilistic load flow is proposed. Corresponding metrics, namely stochastic electrical betweenness and stochastic net-ability are proposed respectively and applied to analyze the vulnerability of power network with wind generations. The case study of CIGRE medium voltage benchmark network is performed for illustration and evaluation. Furthermore, a cascading failures model considering the stochastic metrics is also developed to verify the effectiveness of proposed methodology.