Research On Production Scheduling Of Preventive Maintenance Based On Genetic Algorithm

Along with the growing equipment scale of modern enterprises, equipmentmaintenance plays an important role in improving the equipment reliability andutilization of the production line. And it’s difficult to formulate the productionscheduling because of the occurrence of equipment failure. The production plan madewithout considering equipment maintenance, however, can’t run properly in practicaloperation. This paper will study preventive maintenance and production scheduling asa whole, and mainly discuss the following three production scheduling problemsabout preventive maintenance.In the first part, we study the scheduling problem of jobs with release time onparallel machines with preventive maintenance, where the equipment preventivemaintenance cycle interval is decided based on experience and the machineperformance, and the optimization objectives is to minimize the total duration of jobsas. A genetic algorithm is established to solve this problem, where chromosomesrepresent job sequences and machine sequences, and a first adaptation heuristicmethod is designed as decoding of genetic algorithm to determine the optimalschedule of the job. Numerical experiments show that the genetic algorithm designedin this paper can get better solutions for this problem.In the second part, we study the problem for solving distinct due dates andgeneral early-tardy penalty weights scheduling on parallel machines with preventivemaintenance. Firstly we study the problem of using Weibull distribution to determinethe preventive maintenance cycle interval of equipment, redefining the processingtime of work piece according to the dynamic programming algorithm, and designingthe optimal starting time algorithm to determine the start time and the optimalschedule of the work piece, which are used as decoding of genetic algorithm.Experiment results show that the proposed algorithm is effective.Problem of equipment preventive maintenance of flow shop is discussed in thelast part, we first determine the preventive maintenance cycle model according toWeibull distribution, then use the randomly NEH algorithm as the decoding of thegenetic algorithm to determine the priority sequence and the completion time of jobs.Experimental results show that the proposed algorithm can derive better solution forthese problems.