Abstract
The Brain Storming Optimization (BSO) algorithm is a novel swarm intelligent algorithm that simulates the brainstorming process of humans. This paper presents the BSO algorithm as a solution to the Flexible Job-Shop Scheduling Problem (FJSSP). In aim to improve the global search of the BSO algorithm, a new updating strategy is proposed to adaptively perform several selection methods and neighborhood structures. Furthermore, BSO algorithm has good ability in exploring the search space by clustering the solutions and searching in each cluster independently, thus leading to slow convergence speed, to enhance the local intensification capability and to overcome the slow convergence of the BSO algorithm, we introduce Late Acceptance Hill Climbing (LAHC) with three neighborhoods to the BSO algorithm. Extensive computational experiments were carried out on four well-known benchmarks for FJSSP, and the performance of the BSO algorithm was compared with that of the proposed algorithm. The results demonstrate that the proposed algorithm outperforms the BSO algorithm. Furthermore, the proposed approach overcomes the best-known algorithms in some datasets and it is comparable with these algorithms in other datasets.