Variable neighborhood search for the second type of two-sided
assembly line balancing problem
Deming Lei
a,
n
, Xiuping Guo
b
a
School of Automation, Wuhan University of Technology, Wuhan 430070, Hubei Province, PR China
b
School of economic and management, Southwest Jiaotong University, Chengdu, Sichuan Province, PR China
article info
Article history:
Received 4 July 2015
Received in revised form
21 February 2016
Accepted 5 March 2016
Available online 16 March 2016
Keywords:
Two-sided assembly line balancing
Variable neighborhood search
Precedence
Two- string representation
abstract
Previous studies of two-sided assembly line balancing problem (TALBP) are mainly about the first type of
the problem.TALBP-II which is to minimize cycle time for a given number of stations is seldom inves-
tigated. In this study an effective variable neighborhood search (VNS) is proposed to solve TALBP-II. A
novel two-string representation is used, which is composed of a precedence- based task string and a side
selection string. New solutions are produced by using a side selection operator and two precedence-
based operators. A novel comparison principle is applied to guarantee the feasibility of the solutions and
approximate the optimal solution. VNS is tested on a number of instances and compared with the
existing methods. The computational results show the promising advantage of VNS on the considered
TALBP-II.
& 2016 Elsevier Ltd. All rights reserved.
1. Introduction
Assembly lines have been widely applied in many production
systems to produce high volume standardized products. Generally,
assembly lines can be categorized into two groups: one-sided or
two- sided. In two-sided line, its left side and right side are used in
parallel and tasks are divided into three groups: L-type, R-type and
E-type. L(R)-type task is performed on L(R) side and E-type tasks
can be performed on either side of the line. The length of two-
sided lines can be shorter than the one-sided ones, as a result,
there are many savings on cost and time such as material handling
cost, workers movement, set up time and the cost of tools and
fixtures et al.
Two-sided assembly line balancing problem (TALBP) is the one
of assigning tasks to stations located on two sides in such a way
that some objectives are optimized subject to some specific
restrictions. TALBP is categorized into two versions (Lee et al. [1]):
type-I and type-II. TALBP-I is composed of assigning tasks to sta-
tions such that the number of stations is minimized for a given
cycle time and TALBP-II is to minimize cycle time for a given
number of stations in the assembly line.
TALBP has been considered extensively since the pioneering
work of Bartholdi [2]. For TALBP-I, a number of results are
obtained. Lee et al. [1] and Kim et al. [3] applied genetic algorithm
(GA) for solving the problem. Baykasoglu and Dereli [4] considered
TALBP-I with zoning constraints and proposed ant-colony-based
heuristic algorithm. Simaria and Vilarinho [5] studied the appli-
cation of an ant colony optimization (ACO) to minimize the
number of stations and other objectives. Özcan and Toklu [6]
presented a tabu search (TS) algorithm for TALBP-I with line effi-
ciency and smoothness index. Özcan and Toklu [7] proposed a
simulated annealing (SA) for mixed-model TALBP-I. Özcan and
Toklu [8,9] provided several goal programming models and a
mixed integer programming model for multi-criteria TALBP-I.
Özbakır and Tapkan [10] developed a bees algorithm to solve
TALBP-I with zoning constraints. Khorasanian et al. [11] proposed a
SA for TALBP-I with the relationships between tasks and three
objectives including tasks consistency. Tuncel and Aydin [12]
presented a teaching-learning based optimization for TALBP-I with
multiple constraints such as synchronization and zoning con-
straints. Kucukkoc and Zhang [13] developed a mathematical
model and presented a agent based ACO for the simultaneous
balancing and sequencing of mixed-model parallel two-sided
lines. Li et al. [14] obtained a mathematical formulation of multi-
objective TALBP-I with multiple constraints and proposed an
improved teaching-learning optimization algorithm. Yuan et al.
[15] presented a hybrid honey bee mating optimization to solve
mixed-model TALBP-I.
Unlike TALBP-I,TALBP-II is seldom investigated. Kim et al. [16]
presented a mathematical model and a GA, which adopt the
strategy of localized evolution and steady-state reproduction to
promote population diversity and search efficiency. Zhang et al.
[17] provided a mathematical model of TALBP-II and presented an
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/caor
Computers & Operations Research
http://dx.doi.org/10.1016/j.cor.2016.03.003
0305-0548/& 2016 Elsevier Ltd. All rights reserved.
n
Corresponding author.
E-mail address: deminglei11@163.com (D. Lei).
Computers & Operations Research 72 (2016) 183–188