Effective power system differentiation planning is crucial for enhancing the resilience of power grid infrastructure and bolstering the ability of power systems to manage blackouts. At the heart of power system differentiation planning lies core backbone grid planning. This study involves modeling core backbone grid planning as a multi-objective 0–1 planning problem, which enables the formulation of a multi-objective function that incorporates various factors such as operational and maintenance costs, the significance of nodes and transmission lines, as well as compliance with connectivity and security operation constraints. Moreover, the basic marine predator algorithm was upgraded into a multi-objective optimization algorithm for core backbone grid planning by implementing file management and enhancing the top predator selection mechanism, which managed to fulfill the multi-objective function optimization standards. The results show that in the IEEE 39-node system, the algorithm successfully forms a core backbone grid comprising 22 nodes and 19 transmission lines, achieving economic feasibility with a node-to-line ratio of 1.158. Similarly, for the IEEE 300-node system, the algorithm constructs a larger core backbone grid consisting of 81 nodes and 80 transmission lines, maintaining economic efficiency with a node-to-line ratio of 1.0125. This expanded grid covers a significant number of critical nodes and transmission lines, ensuring optimal network connectivity. Furthermore, the algorithm’s load satisfaction analysis showcases its ability to effectively balance active and reactive power demands, with maximum outputs meeting the respective load demands.