Rescheduling For Rework Jobs On A Single Machine

Effective production scheduling plan is the core of enterprise production and management requirements and academic research hot problems in the complex manufacturing environment. Workshop production often meet with the interference of emergencies, destroyed the original production scheduling of optimality, couldn’t even make it feasible scheduling. At this point, we should make the necessary response to the interference. We should use rescheduling method to repair the initial scheduling and formulate the optimal or nearly optimal scheduling plan for actual production requirement under meeting the various constraint conditions of workshop production. So it not only has important theoretical, but also has practical significance for the study of rescheduling.There is a certain type of unqualified jobs by returning to the location for a simple repair can become qualified jobs under the single machine environment of discrete manufacturing workshop in practical production. So we should schedule these rework jobs on the basis of the initial schedule and meeting all kinds of constraints. This paper studies the rescheduling for rework jobs on single machine (RRSM). According to the characteristics of RRSM problem, we have designed a basic genetic algorithm, rule-guided genetic algorithm and rule-guided adaptive genetic algorithm to solve the problem. We use a large number of experimental examples to verify the solving performance of the three algorithms.First, when the jobs number is small-scale, we determine the optimal parameter combination of three genetic algorithms by uniform design program and a large number of experimental examples. Then we select different examples parameter combinations by uniform design program and simulate different examples under the optimal parameter combination of three genetic algorithms, when the jobs number is small-scale. Last given a representative example parameter combination, we simulate three kinds of genetic algorithm by increasing the size of the problem. Experimental results show that the quality of solution and the efficiency of rule-guided adaptive genetic algorithm are superior to the first two genetic algorithms.In order to further validation the effectiveness of the rule-guided adaptive genetic algorithm, we compare and analyze the results of rule-guided adaptive genetic algorithm, existing heuristic algorithms and branch and bound algorithm respectively through many examples which are given in the simulation experiments. At the same time, we analyze the stability of rule-guided adaptive genetic algorithm under the number of jobs of different scale. Experimental results show that the effect of rule-guided adaptive genetic algorithm would be better and the stability is good.