Non-dominated sorting binary differential evolution for the multi-objective optimization of cascading failures protection in complex networks

A number of research works have been devoted to the optimization of protection strategies (e.g. transmission line switch off) of critical infrastructures (e.g. power grids, telecommunication networks, computer networks, etc) to avoid cascading failures. This work aims at improving a previous optimization approach proposed by some of the authors [1], based on the modified binary differential evolution (MBDE) algorithm. The improvements are three-fold: 1) in the optimization problem formulation, we introduce a third objective function to minimize the impacts of the switching off operations onto the existing network topology; 2) in the optimization problem formulation, we use the final results of cascades, rather than only a short horizon of one step cascading, to evaluate the effects of the switching off strategies; 3) in the optimization algorithm, the fast non-dominated sorting mechanisms are incorporated into the MBDE algorithm: a new algorithm, namely non-dominated sorting binary differential evolution algorithm (NSBDE) is then proposed. The numerical application to the topological structure of the 380 kV Italian power transmission network proves the benefits of the improvements.