Solving Scheduling In Pharmaceutical Industry With Genetic Algorithms

Abstract

The paperwork presents results of applying genetic algorithms in Job Shop Scheduling Problems specific to pharmaceutical industry. These problems proved to be complex enough to require special approaches, from the demand forecast stage to the distribution stage. In this study there were applied, then compared, four genetic algorithms; there are two test cases: a real scheduling problem in pharmaceutical production and the test-instance JSSP-type ft10, both solved as uniobjective and multiobjective problems. The results show that the complex JSSPs can be efficiently solved using the NSGA_II algorithms or the elitist genetic algorithm, depending on the specific type of complexity.