This paper aims to develop a hybrid genetic algo-rithm (HGA) to solve the resource constrained assembly line balancing problem (RCALBP) in the sewing line of a footwear manufacturing plant. Sewing, which is the most critical process in footwear manufacturing, has a series of processes, such as punching, trimming, attaching shoelaces. RCALBP in the sewing line considers not only the prece-dence constraints of product assembly but also the resource
constraints, such as operators and equipment. A novel HGA that includes two stages is proposed to optimize the resources in the sewing line. The first stage uses the priority rule- based method (PRBM) to determine the feasible solutions of assigning tasks and machines to workstations. The solu-
tions of PRBM are used to construct the initial population of genetic algorithm (GA) in the second stage. To ensure that the solution of GA is feasible, a two-point-order crossover with the new technique of searching feasible solution pat-terns is proposed. Moreover, the mutation procedure of GA
is modified to avoid the building block from breaking, which may cause unfeasible solutions in RCALBP. A self-tuning method is also applied recursively to exclude unfeasible solutions. The proposed HGA is compared with the man- ual procedure adopted practically in factories, the existing heuristic model in the literature, and the traditional GA. Based on actual data from a footwear factory, computational results demonstrate that the proposed HGA can achieve bet-ter results than the other algorithms.

EB00000002841K

Available