Resilient Resource Allocations for Multi-stage Transportation-Power Distribution System Operations in Hurricanes

Abstract: The resilience of distribution networks (DNs) or transportation networks (TNs) has attracted a wide attention due to the frequent occurrence of extreme natural disasters. However, the existing works usually investigate the resilience of two systems independently, which neglects their coordinated operations. This paper models the interactions of DNs and TNs and coordinates multiple resilience enhancement strategies to make a more cost-effective resource allocation scheme for the operation of transportation-power distribution networks (TDNs) in hurricanes. Specifically, resource allocations and TDN operations are designed as a two-stage mixed-integer stochastic programming (TMISP) model, where the first stage is to make resource allocation decisions before the hurricane strikes, and the second stage is to perform multi-stage resilience enhancement strategies to minimize the expected TDN operational losses. The uncertain TDN outages with defensive resources in hurricanes are modeled as decision-dependent uncertainties, which are decoupled into decision-independent uncertainties for generating TDN outage scenarios. A comprehensive scenario reduction method is utilized to reduce the redundancies in simulated scenarios. The penalty-based Gauss-Seidel approach is combined with a scenario-skip technique to solve the TMISP problem with binary variables in both stages. Numerical simulations show that the proposed solution methods provide superior performances in computational efficiency and accuracy, over the traditional scenario reduction and scenario-wise decomposition algorithms in solving the TMISP problem. The investment efficiency is increased with the coordination of various DN and TN resources and multiple resilience enhancement strategies, compared with those of separate allocation schemes.