A Real-time Two-step Multi-objective Planning Framework for Resilience Improvement of Islanded Microgrids based on MPC

Abstract

Improving the resilience of islanded microgrids is crucial to ensure power supply continuity against high-impact events. In this paper, a novel two-step multi-objective planning framework based on model predictive control is presented for a microgrid real-time energy management system. The first step involves developing a techno-economic model to optimize the operation of microgrids under normal conditions. A resilience-oriented model is employed in the second step when the microgrid operates in islanded mode due to the upstream grid outage. Using the optimal stored energy calculated in the first step increases the adaptability of the microgrid to abnormal conditions to supply demand. Model predictive control is utilized to optimize energy management in real-time using forecasted data and multi-objective control capability considering operational constraints. In each time step of the framework, a real-time metric based on the percentage of supplied demand is used to determine the microgrid operation mode. Simulation results indicate that the proposed method significantly outperforms counterpart MPC-based methods, achieving superior techno-economic performance under normal conditions. During abnormal conditions, the resilience metric in the proposed method is above 80 % in all cases and is more compared to the counterpart. The capability of the microgrid, load shedding, and redundancy are analyzed in three cases to examine robustness, survival, and agility criteria. A quantitative discussion of these criteria and the resilience metric that supports the comparative analysis proves the improvement of microgrid resilience with the proposed framework in all cases.