Determining optimum Genetic Algorithm parameters for scheduling the manufacturing and assembly of complex products

Abstract

A Genetic Algorithm-based Scheduling Tool (GAST) has been developed for the scheduling of complex products with multiple resource constraints and deep product structure. This includes a repair process that identifies and corrects infeasible schedules. The algorithm takes account of the requirement to minimise the penalties due to both the early supply of components and assemblies and the late delivery of final products, whilst simultaneously considering capacity utilisation. The research has used manufacturing data obtained from a capital goods company. The Genetic Algorithm scheduling method produces significantly better delivery performance and resource utilisation than the Company plans. Genetic Algorithm programs include a number of parameters including the probabilities of crossover and mutation, the population size and the number of generations. A factorial experiment has been performed to identify appropriate values for these factors that produce the best results within a given execution time. The overall objective is to use the most efficient Genetic Algorithm parameters that achieve minimum total costs and minimum spread, to solve a very large scheduling problem that is computationally expensive. The results are compared to the corresponding plans produced by the collaborating company using simulation. It is demonstrated that in the case considered, the Genetic Algorithm scheduling method achieves on time delivery and a 63% reduction in costs.