The ever-increasing penetration of Renewable Energy Sources (RESs) into the power system has faced system operators with higher risks subject to a growing level of the associated uncertainties. To preserve the system frequency security, an under-frequency load shedding (UFLS) scheme can usually be utilized as a final remedial action, which is aimed at removing the excessive load. UFLS can be managed in a multi-stage portfolio based on the priority and sensitivity of the loads under control to cope optimally with the occurring power imbalances. Design of an optimal, robust UFLS scheme is a vital procedure. To that end, the present paper proposes a new UFLS protective system which is conducted to shed the minimum optimal load after precise detection of the frequency excursions. The problem is transferred into a Mixed-Integer Linear Programming (MILP) based optimization framework, and it is to be solved in several stochastic scenarios for setting the UFLS system parameters. The analysis of the UFLS setting results, which are extracted through implementation of the model on the IEEE 39-bus test system, demonstrates the effectiveness of the proposed MILP-based methodology in dealing with the severe uncertainties resulting from RES output variations and load fluctuations.
