Equipment manufacturing job shop dynamic scheduling with worker fatigue and learning effect

Abstract

Aiming at the problem of integrating optimization of worker fatigue and learning effects faced by human and machine dual-resource scheduling in large equipment manufacturing workshops, a mixed-integer programming model was constructed based on exponential fatigue characterization and DeJong learning curves. In accordance with the coding characteristics of the dual-resource scheduling problem, an initial scheduling scheme was generated based on sorting rules, and an adaptive large neighborhood search algorithm was designed by combining removal and insertion operations to solve sub-problems, such as the allocation of human and machine resources, job sequencing and dynamic scheduling. A rescheduling strategy was also designed to address disturbances such as urgent job insertions, machine failures and worker absences. Based on the original algorithm, dynamic scheduling was implemented and tested. The results show that compared with solvers, rules, genetic algorithms and other methods, the proposed algorithm can find scheduling solutions with shorter completion times, and it also performs better in dynamic scheduling scenarios such as urgent orders insert, machine breakdowns and worker absences. The proposed equipment manufacturing workshop scheduling problem takes into account worker fatigue and learning effects, and the corresponding algorithm can provide reference for the development of large-scale equipment manufacturing production scheduling systems.

Keywords

• computer aided manufacturing; job shop scheduling; worker fatigue; learning effect; adaptive large neighborhood search; dynamic scheduling