Design of a method for scheduling in process industry

This paper presents a method for scheduling in the process industry that generates feasible and optimized schedules. In detail the method determines an optimized job sequence and an optimized processor reservation, whereby the two scheduling steps are not executed successively but simultaneously. For the purpose of optimization an evolutionary algorithm is selected that is designed in a specific way in order to reach high quality results. Evaluation results show the ability of the present method to calculate schedules that decrease sequence dependent cost significantly, independent from plant structure and job pool size. In comparison to simple heuristics used in practice to solve problems of real life size the developed method reduces the relevant costs significantly. The method is evaluated with the result to be qualified for application in industrial practice