DIFFERENTIAL EVOLUTION FOR SOLVING CONSTRAINED AND LARGE-SCALE OPTIMIZATION PROBLEMS

Abstract

Differential Evolution is a family of nature-inspired metaheuristic algorithms for solving various optimization problems. The most significant advantage of differential evolution is that it does not make any assumptions about the characteristics and the underlying landscapes of the optimization problem being considered. Thus, differential evolution can tackle a variety of optimization problems even when decision variables and objective functions have complex characteristics. Decision variables having complex characteristics are very common in most real applications, but can be hardly handled by conventional optimization algorithms. In this thesis, some new differential evolution algorithms are proposed to handle two representative complex characteristics of decision variables, and are applied to solve real problems.