Two-stage mobile emergency generator dispatch for sequential service restoration of microgrids in extreme conditions

Abstract

A fast and reliable sequential service restoration (SSR) methodology is necessary to enhance the resilience of the distribution system (DS) after an outage. However, traditional SSR strategies ignore the utilization of mobile emergency generators (MEGs) to reduce the system outage duration. This paper exploits the enhanced resilience of DSs by dispatching MEGs as backup sources to form microgrids (MGs) and sequentially restore out-of-service loads. A two-stage dispatching model is proposed for the SSR of MGs that takes into consideration the preventive control stage (PCS) and the emergency control stage (ECS). Moreover, SSR decisions are optimized depending on stage-based uncertainties. Specifically, in PCS, MEGs are pre-positioned prior to event strikes. Considering contingency uncertainty, a scenario-based stochastic model is proposed to minimize the expected outage duration of loads, and a progressive hedging algorithm is applied to improve computational efficiency. With the realization of contingency in ECS, MEGs are rerouted from pre-positions to target locations, and provisional MGs are sequentially formed to restore critical loads. Considering wind power uncertainty, a robust model is formulated, and the convex hull relaxation method is customized to reduce the computational complexity. The resulting SSR scheme can be adaptively and securely adjusted in response to random contingencies and wind power fluctuation. The effectiveness of the proposed method is validated by numerical simulations.