Generating switching sequences-a genetic algorithm approach

Research for a reliable solution to the problem of generating power system switching sequences has been in progress since the 1960s. But a satisfactory solution cannot be achieved by numerical methods. Recently, some artificial intelligence techniques were applied. In this paper, genetic algorithms are studied. A hybrid system was developed in which a modified simple genetic algorithm (SGA) searches the problem space and an intelligent simulation algorithm qualifies the solutions found. To be able to deal with sequences of different length, heterozygous population was used and the SGA had to be modified. The paper presents this modifications, the encoding and qualifying processes and gives the results which proved the cleverness of the proposed approach