Assessment of makespan performance for flexible process plans in job shop scheduling

摘要

This paper presents a simulation-based Genetic Algorithm (GA) approach to solve Flexible Job Shop Scheduling and assess the effect of flexible process plan of a part-type in a production order on makespan. Simulation provides a meaningful understanding of real phenomenon of a system nature. A regeneration scheme as suggested in literature is also embedded into GA to prevent premature convergence. An algorithm is developed to select a key part-type in order to change the flexible process plan. Two case studies of varying sizes have been considered to assess the performance of flexible job shop with an objective to minimize makespan. Results indicate that by using the flexible process plan of a part-type in a production order assists in reducing the makespan of a production order. Also, selecting the best process plan among flexible process plans on the basis of minimum total production time criterion for a part-type may not yield optimal schedule.