A heuristic algorithm for multi-criteria dynamic scheduling of non-identical parallel machines

Development of technology and customer needs persuade firms to establish make-to-order production systems. Competitiveness makes a necessity of firms to be more reactive regarding the needs of market. Scheduling systems must be developed properly, considering in dynamic conditions of production to provide reactiveness. Changes of complexity and dynamic conditions in production systems have made necessary to develop new scheduling algorithms. Mathematical models are inadequate in solving complex scheduling problems. In this thesis, a specific production scheduling problem which contains non-identical parallel machines, sequence dependent setup times, priorities of customers, multi constraints and multi objectives regarding dynamic production environment is handled. A heuristic production scheduling method is proposed using an algorithmic approach with priorities of customers and time windows of orders as new additional constraints and the model is adapted for an unspecified non-identical parallel machines flexible resource scheduling system. Steps of the proposed heuristic scheduling method are given together with pseudo-code and flow diagram. In this thesis, a dye house department of a textile firm is selected for case study application. The algorithm is adapted regarding the specialties of dye house scheduling problem. The proposed heuristic scheduling algorithm is aimed to increase usage ratio of machine capacity and decrease total setup time, work-in-progress inventory level, lateness of orders, and total completion time. More satisfaction of customers and revenue are expected. As result, in comparison with EDD and SPT rules which are appropriate to use for case study firm, it can be said that the developed heuristic algorithm indicates better performance.