Rolling self-evolution of an aero-engine assembly shop in uncertain knowledgeable manufacturing environment

Abstract

For an aero-engine assembly shop with uncertain number of assemblies and sequence-dependent setup times of work groups, the self-evolution problem of knowledgeable manufacturing systems (KMS) in uncertain manufacturing environment is studied. Rolling horizon method is combined with the hybrid event-driven and periodic-driven self-evolution mechanism to implement the self-evolution of the aero-engine assembly shop, and a feasible rolling rule is given. A mathematical model of the static decision sub-problem at each decision moment is established. For the model, a two-level genetic algorithm is proposed to solve it. In the lower-level hybrid GA, a direct decoding algorithm is given, and the variable neighborhood search algorithm is introduced to enhance the ability of local search. The performance of the algorithm is tested by simulations. In addition, numerical experiments show that the system with self-evolution operations has a better production performance. Self-evolution plays a much larger role especially for the system which is particularly sensitive to the self-adjusting.