Pre-disaster allocation and post-disaster dispatch strategies of power emergency resources for resilience enhancement of distribution networks

•During the pre-disaster prevention period, considering uncertainties in system damage and the material demands for FCs, along with decision-makers' risk preferences, a two-stage DRO model based on mean-CVaR to optimize the allocation of PRMs is established.•During the post-disaster restoration period, based on the results of the pre-disaster PRMs layout, a multi-source cooperative spatio-temporal dispatch model with the primary objective of minimizing power outage losses is formulated, which also considers the dual-coupling relationship between the dispatch of PERs and RCs.•In order to validate the necessary to consider the limited supply of PRMs and the risk-averse attitude of decision makers in enhancing DNs resilience, five control groups are established following existing research methods for comparative analysis. With the frequent occurrence of large-scale power outages caused by extreme disasters, coordinated dispatch of power emergency resources (PERs) and repair crews (RCs) has gradually become an essential method to enhance the resilience of distribution networks (DNs). Previous works solve this coordinated dispatch problem in a risk-neutral approach with the assumption that the supply of power repair materials (PRMs) is still fully available after the disaster. However, growing evidence indicates that certain extreme disasters may lead to limited availability of PERs in the DN, rendering risk-neutral decision making impractical. To fill this gap, this paper proposes a resilience-oriented PERs pre-disaster allocation and post-disaster dispatch strategy in a risk-averse manner. In the pre-disaster prevention process, a two-stage distributionally robust optimization (DRO) model based on mean-CVaR is developed, considering the uncertainties of both system damage and PERs demands, along with the decision maker's preference for risk. In the post-disaster restoration process, a collaborative spatio-temporal dispatching model for PERs and RCs with the goal of minimizing outage losses is established. To tackle the computational challenges associated with PRMs allocation based on DRO, robust counterpart transformation and dual theory are utilized. Case studies are conducted using a real world modified 118-bus distribution network. In comparative analysis, five control groups are established to validate the necessity to consider the limited supply of PRMs and the risk-averse attitude of decision makers in enhancing DNs resilience.