SCHEDULING OF BATCH PROCESSES

KAMAL KURIYAN, Purdue University

Abstract

Scheduling problems occur whenever a processing system is used to produce multiple products. In this context scheduling consists of determining the order in which the different products utilize the resources of the system and the intervals of time over which this utilization is to occur. The objective is to optimize an appropriate measure of system performance. In this work optimization algorithms are developed for the network flowshop scheduling model. This model postulates a general serial-parallel network of processors and is therefore particularly appropriate for the analysis of scheduling problems arising in the CPI. The objective is to develop algorithms which can generate sub-optimal schedules in an efficient manner. Research in this area indicates that all algorithms that generate optimal schedules will probably have worst case running times that increase exponentially with the number of products, and therefore cannot be used on some of the larger problems that are encountered in practice. The initial focus of this work is on a special case: the parallel flowshop problem. Extended versions of list scheduling and MULTIFIT are used to solve this problem. The performance of these heuristics is evaluated by means of computer simulation. List scheduling is shown to be an effective approach when it is coupled with a local search procedure. A rigorous worst case analysis is presented for selected special cases. Heuristics for the general problem are presented in the latter half of the thesis. The performance of these heuristics is also evaluated by computer simulation. Local search procedures are once again found to be a viable means for constructing good sub-optimal schedules. This section is followed by an account of a case study in which heuristics were used to generate schedules for a dairy plant. The final section identifies areas where further work is required.

Degree

Ph.D.

Subject Area

Chemical engineering

COinS