In order to improve the ability to protect the important load under the fault status, it presents a construction method based on the different needs of the core backbone grid. The method, respectively, establishes the model of three core backbone grid from loads, sources and grids needs, and improved the network connectivity-repairing strategy. Improved quantum binary particle swarm optimization (IQBPSO) with dynamic rotation angle, premature judgment mechanism and chaotic mutation strategy is applied to solve the model, and search results are compared and analyzed with quantum binary particle swarm algorithm (QBPSO) and quantum-inspired evolutionary algorithm (QEA). It is applied to the IEEE-118 system, and the results show that the method can accurately search out the core backbone network based on loads, sources and grids needs, and the proposed improved quantum particle swarm algorithm has fast convergence and better optimum result and can overcome the local optimum.