The ordinal optimisation of genetic control parameters for flow shop scheduling

Wang, L.; Zhang, L.; Zheng, Da-Zhong

doi:10.1007/s00170-002-1509-6

Cited by 5 publications

(1 citation statement)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Grefenstette [18] formulated the problem of parameter tuning as an optimisation problem and applied a meta-GA to determine the optimal parameter combination. Wang et al [19] formulated such a problem as a stochastic optimisation problem and developed an ordinal optimisation-based determination method. On the other hand, parameter control is used to dynamically control the parameters during the searching process.…”

Section: Introductionmentioning

confidence: 99%

An adaptive genetic algorithm with multiple operators for flowshop scheduling

Zhang

Zheng

2005

Int J Adv Manuf Technol

View full text Add to dashboard Cite

Genetic algorithms (GAs) are a class of effective parallel searching algorithms inspired by the idea of "survival of the fittest", which has been successfully applied to a variety of problems, especially in the fields of manufacturing and scheduling. However, it is reported that traditional GAs often suffer from the weaknesses of premature convergence as well as parameter and operator dependence. So far, many improved methods with adaptive parameters or hybrid structures have been proposed, but there is little literature considering the adaptive control of genetic operators. In this paper, an adaptive GA (AGA) with multiple operators is proposed for flowshop scheduling, which is a typical NP-hard optimisation problem with many industrial applications and has been widely studied in both academic and engineering fields. In AGA, multiple different genetic operators are employed in an adaptive hybrid way to enhance the exploration and exploitation abilities so as to prevent premature convergence and achieve superior performance. It especially important to stress that the utilising ratio of each operator for hybridisation is adaptively and dynamically controlled during the evolutionary searching process. Simulation results based on benchmarks demonstrate the effectiveness of AGA by contrast with traditional GAs. And the effect of the adaptive control of the operator and the effects of some parameters on the optimisation performance are discussed as well. KeywordsAdaptive control of operator · Flowshop scheduling · Genetic algorithm · Multiple operators Notation n Number of jobs m Number of machines Processing time of job i on machine j C max , C * Makespan, optimal makespan value or lower bound P s Population size G max Maximum generation number δ Basic utilising ratio of operators T Utilising times of an operator N c Number of candidate crossover operators N m Number of candidate mutation operators c c− j , r c− j Contribution and utilising ratio of the jth crossover operator c m−k , r m−k Contribution and utilising ratio of the kth mutation operator a[], b[] Parent individuals c[], d[] Offspring individuals d i Paired difference between the solutions of two algorithms d Mean value of d i s d

show abstract

Section: Introductionmentioning

confidence: 99%