Tardiness heuristic for scheduling Flexible Manufacturing Systems

Abstract

Abstract. This paper evaluates the tardiness performance of a complex dynamic manufacturing environment. These sampling-based adaptive heuristic in a dynamic manufacturing 'changes' include machine breakdowns, unavailability of environment. A test bed, following a real world manufacturing material and other resources to perform manufacturing system, has been developed. The proposed algorithm has been operations, changes in demand, customer priorities and implemented in this simulated cnvironmenL After fme tuning the algorithm, it has been tested in various shop conditions. required delivery dates. To be able to respond to theseThe results of these simulation studies arc summarized.changes effectively, a scheduling system must be able to develop new feasible schedules in a short time. Further more, the proposed schedule should provide a reasonably 1. Introduction good performance of the system for the selected criteria. Our experience with the feedback heuristic (Kiran and The effective utilization of manufacturing systems Alptekin, 1989) indicates that the feedback heuristic requires efflcient scheduling and control systems. Such is a viable alternative for the scheduling of flexible systems must be capable of responding to changes in a manufacturing systems.

Design and planning problems in flexible manufacturing systems: a critical review

“…A more efficient branch and bound algorithm for a similar formulation to the loading problem is presented in Kim and Yano (1991b). The loading problem has also been modelled as part of a general mathematical program for the FMS set-up problem in Kiran and Tansel (1985a), and was shown to be NP-complete by Kiran (1986).…”

“…Models addressing the interaction of planning subproblems have been developed (Kiran andTansel, 1985a andRajagopalan, 1986). The resulting mathematical programming formulations face the curse of combinatorial complexity.…”

“…Given the product mix and the unit operations required by each part type, the decision faced is that of allocating the limited number of fixtures to parts in such a way that all required operations can be performed and an operational objective (usually throughput) is optimized. The fixture allocation problem was addressed as part of a more general model in Kiran and Tansel (1985a). The structure of the formulation bears significant resemblance to the operations assignment problem (refer Section 3.3).…”

“…They use the unit operation as the basic unit of analysis of parts operations. Kiran and Tansel (1985a) have given a precise definition of a unit operation as:…”

“…In addition, the constraint for the non-violation of time availability at each machine is used. Kiran and Tansel (1985a) realized the importance of the congestion phenomena in a FMS and tried to introduce constraints to avoid them. Their modelling approach is indirect.…”