The Research Of The Flexible Job-shop Schedule With Lot-splitting

Job-shop scheduling is important content of workshop production management in manufacturing industry. It considers material, equipment, personnel and other constraints to manage products processed on machine in order to optimize the goal. The flexible job-shop scheduling with lot-splitting as the closest to the actual job-shop scheduling model at this stage, it becomes the focus of people’s pursuit. The batches of workpieces and the management of processes after batching are the most important module of flexible job-shop scheduling with lot-splitting problem. Based on these two problems and get the optimal solution, this thesis mainly includes the followings:(1)Focus on the merit of process parallelism existing in the actual production, we further proposed a new scheduling model based on—The flexible job-shop batch scheduling model for parallel processes. According to the different batch object, we divide model into batch scheduling based on workpieces and processes. Then analyzed the characteristics of these two batch program, then proposed a different solution.(2)For the batches randomness in the model of batch scheduling based on the workpieces, a testing and feedback method to batch different artifacts is proposed. This method make the batch directionally and the search space predictability. After the batching, due to the parallel processes between workpiece increased, so in the solution of genetic algorithm, the decoding method which solves the process parallel is proposed. It shortens the time of the workshop and shop floor management and workshop management complexity.(3) For the process schedule abstract to the workpiece in the mass customization workshop and the inflexibility batch schedule base on the workpiece, we propose a new flexible job shop batch schedule base on the processes. It is multi-processes abstract to the workpiece processed in batch and synthesis. At first, For the process batch problem, we proposed partial solution based on critical path, then dynamic update the process assembly and use the machine selection strategy to determine scheduling processes, at last, get the batching program and process scheduling.