A multi-uncertainty-set based two-stage robust optimization to defender-attacker-defender model for power system protection

•A defender-attacker-defender problem is set up with multiple uncertainty sets.•The probabilities of uncertainty sets are obtained by analytic hierarchy process.•The effect of load types on the proposed model is investigated.•The proposed model is solved by the column-and-constrains generation algorithm.

To handle rapidly growing threats from deliberate attacks, the critical components in power grid should be identified and protected. This paper proposed a defender-attacker-defender model to deal with power grid protection problem, in which the uncertain attacks and load types are considered. Furthermore, multiple uncertainty sets are introduced to characterize the possible realizations of disruptions caused by attackers, and the probabilities of the multiple uncertainty sets are estimated using analytic hierarchy process. Then, the problem is formulated as a multi-uncertainty-set based two-stage robust optimization model which can be termed as a mixed-integer tri-level programming and solved by column-and-constrains generation algorithm with a master-subproblem framework. The test results on a standard IEEE RTS 24-bus system show the effectiveness of the proposed model by considering multiple uncertainty sets and load types.