In order to further describe the virus propagation of real-life complex networks, this paper improves the traditional construction methods of weighted scale-free networks model which considers two key factors:flow bandwidth and individual resistance. Using mean-field theory to simulate the process of the virus transmission, this article analyses the experimental data and verifies the validity of the new model. Most real-life complex networks are known to us with only the local topology information and the traditional virus immunization strategies are based on global network topology information. In condition of knowing local topology information, this paper proposes the immunization strategy of virus spreading based on the local optimum in weighted scale-free networks. Compared with the random immunization strategy and target immunization strategy about the efficiency of virus spreading in weighted scale-free networks, the local optimum immunization strategy is verified to be valid through the dynamic simulation of virus propagation.