The increasing occurrence of destructive natural disasters, such as floods and earthquakes, has posed significant challenges to modern power sys-tems. These events, which are characterized as low-probability but high-impact disasters, impose substantial damage on power systems by causing widespread power outages. Therefore, studying these events and evaluat-ing their impacts on power systems has become increasingly important under the concept of power system resilience, which refers to the ability of a system to withstand adverse conditions and recover rapidly to its normal operating state. In this paper, a resilience model is proposed for a power distribution network containing a network of interconnected mi-crogrids under major disruptive events, considering both centralized and decentralized management approaches. Within the proposed resilience framework, the primary objective is to minimize the load shedding caused by large-scale disturbances within the microgrids while simultaneously minimizing the value of lost load (VOLL) across the entire system, in-cluding both the distribution network and the microgrids. In other words, this paper presents an optimal load-shedding strategy through microgrid interconnection and distribution network reconfiguration while consider-ing the energy constraints of distributed generation units. The optimiza-tion problem is solved under both centralized and decentralized manage-ment schemes. Another objective of this study is to evaluate and compare the performance of the centralized and decentralized management ap-proaches from different perspectives.The proposed model is validated on the modified IEEE 123-bus distribution test system. The developed for-mulation is a Mixed-Integer Linear Programming (MILP) model and has been implemented and solved using the GAMS optimization environment.