Abstract: Based on the Barabasi-Albert scale-free network model and the local-world evolving network model,a hybrid network model is proposed,which is between the two models.Networks can be used to describe the topology systems,in which the nodes represent the system elements and the edges represent the interactions between them.During the growing process of this hybrid model,the nodes added in every time-step are based on their own internal characteristics or external conditions,and optimally connect with the pre-existing nodes in either the local or global preferences,thus creating a network with different local topologies.By adjusting the parameter p,the model can be freely transformed between the Barabasi-Albert scale-free network model and the local-world evolving network model.Simulations indicate that the proposed hybrid model can no only maintain the network robustness of the Barabasi-Albert scale-free network model and the local-world evolving network mode,but reduce their fragility against intentional attacks as well.