Two-level imperialist competitive algorithm for energy-efficient hybrid flow
shop scheduling problem with relative importance of objectives
Ming Li, Deming Lei
*
, Jingcao Cai
School of Automation, Wuhan University of Technology, Wuhan City 430070, China
ARTICLE INFO
Keywords:
Hybrid flow shop scheduling
Imperialist competitive algorithm
Two-level
Energy-efficient
Relative importance of objectives
ABSTRACT
Energy-efficient hybrid flow shop scheduling problem (EHFSP) has been investigated in recent years; however,
the relative importance of objectives is seldom considered in the previous works. In this study, EHFSP with total
tardiness, makespan and total energy consumption is addressed, in which the third objective has lower impor-
tance than other ones. A new Pareto dominance is defined to deal with the relative importance and a two-level
imperialist competitive algorithm (TICA) is presented, in which two levels consist of the strongest empire and
other empires, respectively. To generate high quality solutions, assimilation and revolution are executed differ-
ently in empires in the different search stages, the strongest empire is excluded from imperialist competition,
memory is often combined with the strongest empire and a member of memory is added into the winning empire
to avoid the inclusion of the weakest colony of the weakest empire. Extensive experiments are conducted and the
computational results show that TICA provides promising results for the considered EHFSP.
1. Introduction
Hybrid flow shop scheduling problems (HFSP) are not uncommon in
many real-life manufacturing industries such as electronics, paper,
textile, petrochemical, airplane engine and semiconductor [1,2].
Compared with flow shop, hybrid flow shop exists more than one parallel
machine at one stage at least and the redundancy of machines can result
in flexibility and the increasing capacities and avoid bottleneck etc [3]. In
the past decades various scheduling problems in hybrid flow shop such as
multi-objective HFSP and energy-efficient hybrid flow shop scheduling
problem (EHFSP) have been addressed.
With respect to multi-objective HFSP, Jungwattanakit et al. [4] pro-
posed some heuristics and a genetic algorithm (GA) for the problem with
unrelated machines, setup times and dual criteria. Rashidi et al. [5]
proposed an improved hybrid parallel GA. Cho et al. [6] reported a
parallel GA with four different versions of local search strategies for
reentrant HFSP. Karimi et al. [7] developed a multi-phase GA for
bi-objective hybrid flexible flowshop group scheduling problem. Naderi
et al. [8] solved HFSP with sequence-dependent setup times, trans-
portation times and two objectives using an improved simulated
annealing. Tran and Ng [9] applied a hybrid water flow algorithm for
HFSP with limited buffers. Mousavi et al. [10] presented a bi-objective
local search algorithm with three phases. The previous works also
considered HFSP with other practical constraints such as preventive
maintenance [11], two agents [12], family setup times [13] and assembly
[14]. Other meta-heuristics are also applied, which include tabu search
[11,15], colonial competitive algorithm (CCA) [16], neighborhood
search [14,17], variable neighborhood search (VNS) [18], shuffled
frog-leaping algorithm [12] and firefly algorithm [19] etc.
EHFSP is an important energy-efficient scheduling problem which is
the important part of green manufacturing, and often has multiple ob-
jectives because of at least one objective or constraint is related to energy
efficiency. In recent years, EHFSP has attracted much attention. Bruzzone
et al. [20] proposed a method to modify the scheduling of the jobs in
order to account for power's peak. Dai et al. [21] developed a
genetic-simulated annealing algorithm to minimize makespan and total
energy consumption in flexible flow shop scheduling. Luo et al. [22]
presented a novel ant colony optimization for HFSP with the consider-
ation on electricity consumption cost. Tang et al. [23] gave an improved
particle swarm optimization (PSO) for energy-efficient dynamic sched-
uling in flexible flow shop. Lin et al. [24] presented
teaching-learning-based optimization (TLBO) for the integration of pro-
cessing parameter optimization and HFSP with the minimization of
makespan and carbon footprint. Yan et al. [25] proposed a multi-level
optimization method for reducing the total energy consumption and
makespan. Lei et al. [26] developed a novel TLBO to minimize total
* Corresponding author.
E-mail address: deminglei11@163.com (D. Lei).
Contents lists available at ScienceDirect
Swarm and Evolutionary Computation
journal homepage: www.elsevier.com/locate/swevo
https://doi.org/10.1016/j.swevo.2019.05.006
Received 11 December 2018; Received in revised form 12 April 2019; Accepted 20 May 2019
Available online 30 May 2019
2210-6502/© 2019 Published by Elsevier B.V.
Swarm and Evolutionary Computation 49 (2019) 34–43