|
Abstract
|
A huge number of sophisticated optimization problems have recently emerged in numerous disciplines of engineering, science, economics, and so on. The search space for these problems is enormous, and traditional optimization approaches cannot identify optimal solutions. Stochastic optimization methods, such as meta-heuristic algorithms, have been successfully applied to a wide range of modern and critical fields, including pollution routing, mobile robot navigation, system identification, artificial intelligence planning problems, transformer diagnosis, communication applications, image processing applications, safety analysis problems, and many more. Meta-heuristic optimization algorithms are more popular than classical algorithms because they have several advantages, including the following: (1) Gradient-free: unlike classical optimization methods, they do not employ any complex mathematical processing. (2) Problem independent: they are not dependent on a specific problem and can be applied to any problem. (3) No local optima: because of their stochastic nature, they are able to avoid local optima. (4) Motivated by simple ideas: they are inspired by basic phenomena that do not require the application of complicated mathematical principles. In addition to these characteristics, they have two main characteristics: exploration and exploitation. The first relies on the algorithm's ability to find new solutions in new search spaces, while the second implies that the algorithm should use previously generated solutions to improve the quality of new solutions. Finding a good trade-off between these features is a necessary prerequisite for a meta-heuristic algorithm to be more effective. So far, the city council’s evolution (CCE) algorithm has not been used along with the firework algorithm. As a result, these two algorithms can be combined to improve the performance of CCE and FWA.
|