The Ant Colony Method
摘要
This chapter presents a metaheuristic based on the behavior of ants, illustrating the importance of Nature-inspired methods in this field. Ants are collectively able to perform optimization tasks, despite a lack of central control. The chapter starts by reviewing results based on real ants, so as to highlight the fundamental mechanisms that make ants collaborate and find the optimal way from their nest to a food source. This idea has been translated into a computer algorithms to give rise to the so-called Ant System (AS) and Ant Colony Systems (ACS), two population metaheuristics. These algorithms are presented in their historical form for solving the Traveling Salesman Problem (TSP), but also in a version that optimizes an arbitrary fitness function. This enlarges the scope of the method and allows a simple discussion of its performance.