A Workstation Solution Based Heuristic Algorithm for Assembly Line Balancing Problem

In the automotive industry, assembly line balancing work plays an important role in assembly process planning. Assembly line balancing problem is a kind of NP hard problem, especially for Two-sided assembly line balancing problem. It is too difficult to obtain the optimal result using the commercial solvers directly. This paper presents a heuristic algorithm to deal with the two-sided assembly line balancing problem. Based on the mathematical model of the two-sided assembly line balancing problem, we analyze the characteristics of the problem. A workstation oriented strategy is used to generate assignment for each mated-station. Priority rules are created in the process of solution generation. We implement numerical experiments based on the benchmark data sets. Results verified the effectiveness and efficiencies of the proposed algorithm.

1 Introduction

The automotive manufacturing contains four main processes: Pressing, Joining, Painting and Assembly. Assembly production is the last stage in which all parts and components are installed on the vehicle body. These works are done on the assembly line that was firstly utilized in auto industry by Henry Ford. After a century, production technologies improve, and the assembly line is widely used in many manufacturing situations. The assembly line in a vehicle factory usually consists of a main assembly line and several pre-assembly lines. There are 1500 to 2000 components that need to be installed in the assembly line for a modern passenger vehicle. Planners in the assembly factory design the sequence of installations before the Start of Production (SOP). The assembly line balancing schedule is planned based on these works. Most line balancing works are done manually according to planners personal experiences in the factory. However, as the models increase it is challenging to finish the line balancing work. The assembly line balancing problem is a demonstrated NP-hard problem. An efficient algorithm for the line balancing problem is necessary to generate a high-quality solution automatically.