The Study And Application Of Flow Shop Scheduling Problem Based On Genetic Algorithm

With the increasing of production s scale, the management and monitoring to theproduction process are put forward higher requirements. In order to gain an advantage in thefierce market competition, enterprises adapt advanced scheduling techniques for reasonablescientific production scheduling solutions to improve the efficiency of the entiremanufacturing system of enterprise. Thus, shop scheduling problems have emerged. It stemsfrom a lot of different manufacturing areas and is taken as an abstract model of the actualnumber of production lines and is suitable for manufacturers with background of massproduction of a single piece. It is of great guidance to practical application. In this paper, flowshop scheduling problem with fuzzy delivery time and hybrid flow shop scheduling problemare studied and some certain results has been obtained.Firstly, studies to the background and significance of the subject are introduced briefly.An overview for the general flow shop scheduling problems and solutions to the problems ispresented. The genetic algorithm theory is introduced in detail and the application for geneticalgorithm in the ordinary flow shop scheduling problem is focused and the design of GAoperators, settings for control and other parameters are included.Secondly, the intelligent optimization algorithms are studied in this paper. Theshortcomings of genetic algorithm for flow shop scheduling problem with fuzzy delivery timewere analyzed, namely, GA is easy to fall into local optimum and has the disadvantage of slowconvergence, which makes globally optimal solution can t be obtained. The population isdivided into several sub-populations and chaotic migration strategy is introduced, soindividuals between the several sub-populations can get full exchange of information toincrease the diversity of individuals in the population. A standard GA is combined with NEH local optimization algorithm to achieve a hybrid genetic algorithm (CMGA). Results forstudies verify the effectiveness and rapidity. Finally, the hybrid flow shop schedulingproblems were studied. Estimation of distribution algorithm is introduced for the disadvantageof large computation for solving hybrid flow shop scheduling problem based on GA. Based onmechanism of quality evaluation, individuals generated by such as crossover and mutationoperators are evaluated and select individuals in good quality as offspring, avoiding thecomputational burden caused by calculating the fitness of individuals generated by crossoperators, mutation operators. A self-guided genetic algorithm is designed to verify therobustness and good ability.