A hierarchical parallel multi-station assembly sequence planning method based on GA-DFLA

Abstract

In recent years, the parallel assembly sequence planning (PASP) is put forward to accommodate the development of greater variety in the mass-customization production. Although many researchers have made contributions to assembly sequence planning (ASP), how to construct parallel multi-station ASP model to effectively solve the integrated ASP and assembly line balancing (ALB) problem need to be further investigated. In this paper, a hierarchical parallel multi-station assembly sequence planning method based on the genetic algorithm and discrete frog leaping algorithm (GA-DFLA) is proposed to solve the integrated parallel multi-station ASP and ALB problem. The assembly information datasets of the model based definition (MBD) including parts information, hierarchical information, matrix information and resource information are defined firstly. Then the hierarchical structure tree is constructed according to the divided assembly units. The hierarchical parallel multi-station assembly sequence planning is carried out from bottom to top in hierarchical structure tree model using the GA-DFLA. The fitness function with feasibility index, time cost index and assembly line balance index, is proposed to determine a more appropriate solution. Finally, the rear independent suspension is taken as an example to validate this method. The results show that the time of parallel multi-station model is at least 35.27% less than the time of serial multi-station model.