Hybrid Estimation of Distribution Based on Knowledge Transfer for Flexible Job-shop Scheduling Problem

In this work, we introduce a new method called G-EDA for solving flexible job-shop scheduling problems (FJSP). Based on the investigation of current works, it can be found that the solution of FJSP mainly focuses on the intelligent optimization algorithm, such as the genetic algorithm, which obtains new population through crossover and mutation. When searching individuals, this algorithm does not fully use the knowledge hidden in the excellent individuals in the previous generation population, and its performance is poor when facing high-dimensional problems. This paper hopes to find a method to find the distribution law of high-quality individuals in the previous generation population and transfer the knowledge contained in the previous generation population to the next generation to improve the performance of the algorithm. In this paper, we propose a hybrid estimation of distribution algorithm using population grouping mechanism(G-EDA). We conduct numerical simulations based on two sets of international standard examples. And we compare G-EDA with some existing advanced algorithms. The results show that G-EDA is effective and practical in solving multi-objective FJSP.