Flexible jobshop scheduling with multiobjective based on genetic algorithm

FANG Shuiliang, YAO Yanfei, ZHAO Shikui
(Department of Mechanical Engineering, Zhejiang University, Hangzhou 310027, China)

Abstract: For the multiobjective scheduling of the flexible jobshop problem with a nonzero partarrivetime and machineavailabletime, an improved genetic algorithm was proposed. Doublechain structure with machineallocationchain and operationsequencechain was used to code the chromosome; Population was initialized with three methods: uniform design, shortestprocessingtime machine assignment, and random generation. Population selection was performed with tournamentselection, elitistselection, and newbornselection. A hybrid crossover method was proposed, including traditional crossover, machinebottleneck oriented and jobbottleneck oriented crossover. The mutation was performed with an adaptable probability. The decoding process was controlled with heuristic rules. Finally, casestudies based on some typical benchmarkexamples were carried out to evaluate the algorithm. The results show a quicker speed and powerful optimizing capability.
Key words: flexible jobshop scheduling; genetic algorithm; multiobjective optimization; uniform design