Title
An effective parallel evolutionary metaheuristic with its application to three optimization problems
Publication Name
Applied Intelligence
Abstract
This paper presents a parallel evolutionary metaheuristic which includes different threads aimed at balancing exploration versus exploitation. Exploring different areas of the search space independently, each thread also communicates with other threads, and exploits the search space by improving a common high quality solution. The presented metaheuristic has been applied to three famous and hard-to-solve optimization problems, namely the job shop scheduling, the permutation flowshop scheduling, and the quadratic assignment problems. The results of computational experiments indicate that it is effective, versatile and robust, competing with the-state-of-art procedures presented for these three problems. In effect, in terms of solution quality, and average required running time to reach a high quality solution, the procedure outperforms several state-of-the-art procedures on multiple benchmark instances.
Open Access Status
This publication is not available as open access