Elsevier

Omega

Volume 34, Issue 1, January 2006, Pages 81-89
Omega

Total tardiness minimization on unrelated parallel machine scheduling with auxiliary equipment constraints

https://doi.org/10.1016/j.omega.2004.07.023Get rights and content

Abstract

This research deals with scheduling jobs on unrelated parallel machines with auxiliary equipment constraints. Each job has a due date and requires a single operation. A setup for dies is incurred if there is a switch from processing one type of job to another type. For a die type, the number of dies is limited. Due to the attributes of the machines and the fitness of dies to each, the processing time for a job depends on the machine on which the job is processed, each job being restricted to processing on certain machines. In this paper, an effective heuristic based on threshold-accepting methods, tabu lists, and improvement procedures is proposed to minimize total tardiness. An extensive experiment is conducted to evaluate the computational characteristics of the proposed heuristic. Computational experiences demonstrate that the proposed heuristic is capable of obtaining optimal solutions for small-sized problems, and significantly outperforms an ATCS procedure and a simulated annealing method for problems in larger sizes.

Introduction

This paper deals with the problem of scheduling N jobs on M unrelated parallel machines. Each job has a due date and requires a single operation. A setup that includes detaching one die and attaching another from the appropriate die type is incurred if there is a switch from processing one type of job to another type. Due to the attributes of the machines and the fitness of dies to each, the processing time for a job depends on both the job and the machine, each job being restricted to processing on certain machines. Moreover, the detaching (attaching) time may have different values depending on both the die type and the machine on which the die is detached (attached). Such a production environment can be found in injection-molding departments where parallel machines are used to process different components and for which setups are required for auxiliary equipment (e.g., molds). This scheduling problem is also frequently encountered in the die-casting industry, since many different machines are used to produce castings. The dies (or molds) are quite expensive; thus, the number of dies of each type available is limited. Therefore, dies must be treated as restricted secondary resources, the fact of which distinguishes this study from many past studies in unrelated parallel-machine scheduling in which secondary resources are unrestricted.

It is known that this class of scheduling problems is NP-hard [1]. In this paper, an effective heuristic combining the threshold-accepting (TA) method, tabu list techniques from the tabu search method, and designed improvement procedures is proposed to minimize total tardiness. Computational characteristics of the presented heuristic are evaluated in empirical comparisons with an apparent-tardiness-cost-with-setup (ATCS) procedure [2], a basic simulated annealing (SA) method from the literature [3], and optimal solutions. Computational experiences demonstrate that the proposed heuristic significantly outperforms the ATCS procedure and the SA method, and is capable of obtaining optimal solutions for small-sized problems.

The rest of this paper is organized in five sections. The related research is reviewed in Section 2; the TA method and tabu lists are briefly discussed in Section 3; the proposed heuristic is described in Section 4; computational experiments and results are reported in Section 5; conclusions and suggestions for future research are discussed in Section 6.

Section snippets

Related research

Parallel-machine scheduling with tardiness-oriented performance measures has attracted a lot of research attention in recent years. For identical parallel-machine scheduling, Root [4] considered that all jobs have the same due dates and presented a constructive algorithm to minimize the total tardiness. Lawler [5] assumed that all jobs have the same processing times and formulated the total tardiness problem as a transportation problem. Elmaghraby and Park [6] developed a branch-and-bound

TA method and tabu lists

The TA method and tabu lists are briefly described in this section.

Heuristic H1

The proposed heuristic, H1, is described in this section. First, the following notations are defined:

groupa set of jobs that are allocated to the same machine and require the same type of die
jindex for jobs (j=1,2,,N)
mindex for machines (m=1,2,,M)
dindex for die types (d=1,2,,D)
Inner_maxmaximum number of consecutive iterations without improvement to the best known solution by a specific procedure
Outer_maxmaximum number of consecutive iterations without improvement to the best known solution

Computational experiments

In this section, the computational characteristics and effectiveness of H1 are evaluated by using two sets of newly generated test problems, since there are no “standard” or benchmarking test data in the open literature. The first set of test problems, consisting of problems in smaller sizes, is designed mainly for demonstrating that H1 is capable of obtaining optimal solutions. As for the second set of test problems comprising problems in larger sizes, H1 is compared with an ATCS procedure, a

Conclusions and suggestions for future research

This research has dealt with scheduling jobs on unrelated parallel machines with auxiliary equipment constraints. An effective heuristic based on threshold-accepting method, tabu lists, and improvement procedures has been proposed to minimize total tardiness. Two sets of test problems have been generated to evaluate the computational characteristics of the proposed heuristic. Computational results have indicated that the proposed heuristic is capable of obtaining optimal solutions for problems

Acknowledgements

We wish to thank the anonymous referees for improving this paper through their useful comments.

References (31)

  • C.N. Potts et al.

    Decomposition algorithm for the single machine total tardiness problem

    Operations Research Letters

    (1982)
  • C. Koulamas

    The total tardiness problemreview and extensions

    Operations Research

    (1994)
  • Tamaki H, Hasegawa Y, Kozasa J, Araki M. Application of search methods to scheduling problem in plastics forming plant:...
  • J.G. Root

    Scheduling with deadlines and loss functions on k parallel machines

    Management Science

    (1965)
  • E.L. Lawler

    On scheduling problem with deferral costs

    Management Science

    (1964)
  • Cited by (72)

    • Application of a hybrid evolutionary algorithm to resource-constrained parallel machine scheduling with setup times

      2022, Computers and Operations Research
      Citation Excerpt :

      The tabu search reported the best results, including the optimal results for 24 small instances. Considering the same environment, but with the objective of minimizing the total tardiness, Chen and Wu (2006) presented a hybrid heuristic that combines the threshold accepting method and tabu search. A total of two sets of instances proposed in the study were considered in the experiments.

    • Stochastic customer order scheduling on heterogeneous parallel machines with resource allocation consideration

      2021, Computers and Industrial Engineering
      Citation Excerpt :

      Low and Wu (2016) explore the case where job-processing times are controllable through the allocation of a nonrenewable common resource with eligibility constraints. Besides, some studies also consider resource constrained problems (Afzalirad & Shafipour, 2018; Cappadonna, Costa, & Fichera, 2013; Chen, 2005; Chen & Wu, 2006; Edis & Oguz, 2011; Torabi, Sahebjamnia, Mansouri, & Bajestani, 2013) which can be deemed as special cases of resource allocation problems as resources are necessary to perform operations. To the knowledge of the authors, all these studies focus on jobs instead of customer orders, and stochastic environment is scarcely discussed for unrelated parallel machine systems.

    View all citing articles on Scopus
    View full text