Multi-stage hybrid algorithm-enabled optimization of sequence-dependent assembly line configuration for automotive engine

Abstract

Engines are the most expensive and technology-intensive components in automobiles, so an optimized configuration of the automotive engine assembly line (AEAL) is anticipated to improve efficiency and reduce cost. The traditional methods for assembly line configuration can mainly work out a proper machine number, but they generally ignore process sequences that could also influence the buffer cost derived from the assignment of divergence and confluence buffers. Simultaneously, how to reduce the number of variables in the algorithm iteration process to improve computational efficiency is rarely considered in the existing studies. To bridge the gaps, this study proposes a multi-stage hybrid algorithm based on a backtracking searching algorithm (BSA) to realize an effective configuration that can further improve production efficiency and reduce equipment cost for sequence-dependent AEALs. First, an AEAL configuration model is developed to involve machine number and process sequence as decision variables and aims to satisfy multiple objectives concerned with equipment cost and cycle time. Then, a multi-stage hybrid algorithm is proposed to efficiently acquire the optimal solutions to machine number and process sequence in multiple stages that can improve computational efficiency. Finally, the effectiveness and superiority of the proposed method are validated via a case study. The numerical results show that the proposed method can effectively improve production efficiency and reduce equipment cost for sequence-dependent AEALs with a better convergence and diversity performance.