A new continuous model for job-shop scheduling

In this paper arguments are presented for the use of a new continuous model for job-shop scheduling. In this model each job is regarded as infinitely divisible. Formulating the job-shop problem in this way leads to optimal control problems of a type which recent work has shown to be efficiently solvable. Solutions to these continuous problems can be used to generate heuristic loading rules. Some simulation results from small job-shops are presented. These demonstrate that loading rules based on the continuous model perform well in practice.