The recent experiences of extreme weather events highlight the significance of boosting the resilience of distribution systems. In this situation, the resilience of distribution systems planning leads to an efficient solution for protecting the system from these events via line hardening and the installation of distributed generators (DGs). For this aim, this study presents a new two-stage stochastic mixed-integer linear programming model (SMILP) to hedge against natural disaster uncertainty. The first stage involves making investment decisions about line hardening and DG installation. Then, in the second stage, the dynamic microgrids are created according to a master-slave concept with the ability of integrating distributed generators to minimize the cost of loss of load in each uncertain outage scenario. In particular, this paper presents an approach to select the line damage scenarios for the SMILP. In addition, the operational strategies such as load control capability, microgrid formation and network reconfiguration are integrated into the distribution system plans for resilience improvement in both planning and emergency response steps. The simulation results for an IEEE 33-bus test system demonstrate the effectiveness of the proposed model in improving disaster-induced the resilience of distribution systems.

In recent years, severe outages caused by natural disasters such as hurricanes have highlighted the importance of boosting the resilience level of distribution systems. However, due to the uncertain characteristics of natural disasters and loads, there exists a research gap in the selection of optimal planning strategies coupled with provisional microgrid (MG) formation. For this purpose, this study proposes a novel three‐stage stochastic planning model considering the planning step and emergency response step. In the first stage, the decisions on line hardening and Distributed Generation (DG) placement are made with the aim of maximising the distribution system resilience. Then, in the second stage, the line outage uncertainty is imposed via the given scenarios to form the provisional MGs based on a master‐slave control technique. In addition, the non‐anticipativity constraints are presented to guarantee that the MG formation decision is based on the line damage uncertainty. Last, with the realisation of the load demand, the cost of load shedding in each provisional MG is minimised based on a demand‐side management program. The proposed method can consider the step‐by‐step uncertainty realisation that is near to the reality in MG formation strategy. Two standard distribution systems are utilised to validate the correctness and effectiveness of the presented model.