Population-Based Metaheuristics: A Comparative Analysis

To optimally solve hard optimization problems in real life, many methods were designed and tested. The metaheuristics proved to be the generally adequate techniques, while the exact traditional optimization mathematical methods are prohibitively expensive in computational time. The population-based metaheuristics, which manipulate a set of candidate solutions at a time, have advantages over the singlestate methods and therefore are preferred techniques when hard problems are to be solved. Such metaheuristics include Genetic Algorithms, Ant Colony Optimization, Particle Swarm Optimization, Scatter Search and many more methods. In this survey a comparative analysis of the main population-based metaheuristics was accomplished; the focus is on the fundamental properties regarding operational principle, on the adequate problems, the advantages and disadvantages in use. Keywordsmetaheuristic; candidate solution; optimization.