A genetic continuation power flow approach to assess power system vulnerability

In the present work a procedure for the identification of critical multiple contingencies is proposed. Its aim is that of highlighting those multiple outages which are likely to endanger system security especially in the case of faulty operation of the protection system. The proposed procedure exploits the efficient genetic algorithm approach recently proposed by Chu and Beasley to rank the worst combinations of line or generator out-ages. The evaluation of the fitness corresponding to each individual in the population is carried out by a suitable ac power flow computation based on a continuation method. The procedure was successfully tested on a number of test networks as well as on the Italian power system.