Co-evolutionary genetic algorithm for fuzzy flexible job shop scheduling

Fuzzy flexible job shop scheduling problem (FfJSP) is the combination of fuzzy scheduling and flexible scheduling in job shop environment, which is seldom investigated for its high complexity. We developed an effective co-evolutionary genetic algorithm (CGA) for the minimization of fuzzy makespan. In CGA, the chromosome of a novel representation consists of ordered operation list and machine assignment string, a new crossover operator and a modified tournament selection are proposed, and the population of job sequencing and the population of machine assignment independently evolve and cooperate for converging to the best solutions of the problem. CGA is finally applied and compared with other algorithms. Computational results show that CGA outperforms those algorithms compared.

.Figure optionsDownload as PowerPoint slideHighlight
► Fuzzy flexible job shop scheduling is considered.
► A novel co-evolutionary genetic algorithm (CGA) is proposed to minimize fuzzy makespan.
► In CGA, a new crossover and modified tournament selection are used.
► In CGA, two sub-problems are evolved through cooperation on chromosome.
► Computational results show the promising advantage of CGA on the considered problem.