The USE of a genetic algorithm for solving electric engineering problems

摘要

An advantage of gradient methods is that they provide good convergence of the calculation in the region of attraction to an extreme. However, these methods cannot be used to search a global extreme in problems with many local extremes. An advantage of genetic algorithms is that they can be efficiently used to solve non-smooth problems with many local extremes. A drawback of these methods is that they show poor convergence near an extreme. A modified genetic algorithm (belonging to a group of hybridization algorithms) is proposed, which combines the concepts used in gradient and genetic algorithms.