Multi-Time-Step Rolling Optimization Strategy for Post-Disaster Emergency Recovery in Distribution System Based on Model Predictive Control

Abstract

Aimed at improving the resilience of distribution systems and analyzing the influence of uncertainties on the post-disaster emergency recovery process of a power-transportation system, this paper proposes a multi -time-step rolling optimization strategy based on model predictive control (MPC). First, the prediction models for three types of uncertainties: recovery time of faulty equipment, load demand and photovoltaic output are established. When it is assumed that traffic flow and density can meet a specific relationship, the cell transmission model (CTM) can be used to establish the transportation network. The evidence theory is used to model and analyze the two types of uncertain factors that constitute fault recovery time: travel time and repair time. Second, a mixed integer linear programming model of emergency recovery is established, in order to minimize load reduction and restoration resource scheduling cost. Decision variables include repair sequence of faulty equipment and the scheduling plan for restoration resources. Through the rolling optimization and feedback correction process, load demand can track the expected value and economic loss of power outages can be reduced as much as possible. Finally, a case study is used to verify the effectiveness of the method proposed in this paper and validate a multi-time-step rolling optimization strategy has the advantages of updating uncertain information multiple times, which can achieve the goal of planning a shorter path for the repair crew and shortening the total duration of fault recovery.

Keywords

• Uncertainties
• power-transportation system
• model predictive control
• resilience improvement
• emergency recovery