Metaheuristic algorithms

This chapter begins with a brief introduction to the basic ideas and historical development of metaheuristic algorithms. The aim in different periods was essentially the same: developing new metaheuristic algorithms. The main difference is that the focus is different between different algorithms, in the sense that, for example, in some algorithms, the focus is on using populations and mimicking social behavior; in other algorithms, the focus is on adjusting the parameter settings dynamically; and in others, the focus is on applying metaheuristic algorithms to parallel computing systems. A unified framework for metaheuristic algorithms composed of initialization, transition, evaluation, determination, and output operators is then presented to provide an integrated view of different metaheuristic algorithms. A comparison is made to emphasize the distinguishing features between metaheuristic algorithms, exhaustive search, and greedy algorithms.