Comparison of two heuristic approaches for solving the production scheduling problem

Arunas Andziulis, Dale Dzemydiene, Raimundas Steponavičius, Sergej Jakovlev

    Research output: Contribution to journalArticle

    7 Citations (Scopus)

    Abstract

    Production scheduling problems attract a lot of attention among applied scientists and practitioners working in the field of combinatorial optimization and optimization software development since they are encountered in many different manufacturing processes and thus effective solutions to them offer great benefits. In this work, two commonly used heuristic methods for solving production scheduling problems, namely, the Nearest Neighbor (NN) and Ant Colony Optimization (ACO) have been tested on a specific real-life problem and the results discussed. The problem belongs to the class of Asymmetric Travelling Salesman Problems (ATSP), which is known as a hard type problem with no effective solutions for large scale problems available yet. The performances of the Nearest Neighbor algorithm and the Ant Colony Optimization technique were evaluated and compared using two criteria, namely: The minimum value of the objective function achieved and the CPU time it took to find it (including the statistical confidence limits). The conclusions drawn suggest that on one hand the ACO algorithm works better than NN if looking at the achieved minimum values of the objective function. On the other hand, the computational time of the ACO algorithm is slightly longer.

    Original languageEnglish
    Pages (from-to)118-122
    Number of pages5
    JournalInformation Technology and Control
    Volume40
    Issue number2
    Publication statusPublished - 2011

    Fingerprint

    Ant colony optimization
    Scheduling
    Traveling salesman problem
    Heuristic methods
    Combinatorial optimization
    Program processors
    Software engineering

    Keywords

    • Ant colony optimization
    • Asymmetric travelling salesman problem
    • Nearest neighbor
    • Production scheduling
    • Theory of algorithms

    ASJC Scopus subject areas

    • Computer Science Applications
    • Electrical and Electronic Engineering
    • Control and Systems Engineering

    Cite this

    Comparison of two heuristic approaches for solving the production scheduling problem. / Andziulis, Arunas; Dzemydiene, Dale; Steponavičius, Raimundas; Jakovlev, Sergej.

    In: Information Technology and Control, Vol. 40, No. 2, 2011, p. 118-122.

    Research output: Contribution to journalArticle

    Andziulis, Arunas ; Dzemydiene, Dale ; Steponavičius, Raimundas ; Jakovlev, Sergej. / Comparison of two heuristic approaches for solving the production scheduling problem. In: Information Technology and Control. 2011 ; Vol. 40, No. 2. pp. 118-122.
    @article{dc47451e6c574bc988824feba6c36e63,
    title = "Comparison of two heuristic approaches for solving the production scheduling problem",
    abstract = "Production scheduling problems attract a lot of attention among applied scientists and practitioners working in the field of combinatorial optimization and optimization software development since they are encountered in many different manufacturing processes and thus effective solutions to them offer great benefits. In this work, two commonly used heuristic methods for solving production scheduling problems, namely, the Nearest Neighbor (NN) and Ant Colony Optimization (ACO) have been tested on a specific real-life problem and the results discussed. The problem belongs to the class of Asymmetric Travelling Salesman Problems (ATSP), which is known as a hard type problem with no effective solutions for large scale problems available yet. The performances of the Nearest Neighbor algorithm and the Ant Colony Optimization technique were evaluated and compared using two criteria, namely: The minimum value of the objective function achieved and the CPU time it took to find it (including the statistical confidence limits). The conclusions drawn suggest that on one hand the ACO algorithm works better than NN if looking at the achieved minimum values of the objective function. On the other hand, the computational time of the ACO algorithm is slightly longer.",
    keywords = "Ant colony optimization, Asymmetric travelling salesman problem, Nearest neighbor, Production scheduling, Theory of algorithms",
    author = "Arunas Andziulis and Dale Dzemydiene and Raimundas Steponavičius and Sergej Jakovlev",
    year = "2011",
    language = "English",
    volume = "40",
    pages = "118--122",
    journal = "Information Technology and Control",
    issn = "1392-124X",
    publisher = "Kauno Technologijos Universitetas",
    number = "2",

    }

    TY - JOUR

    T1 - Comparison of two heuristic approaches for solving the production scheduling problem

    AU - Andziulis, Arunas

    AU - Dzemydiene, Dale

    AU - Steponavičius, Raimundas

    AU - Jakovlev, Sergej

    PY - 2011

    Y1 - 2011

    N2 - Production scheduling problems attract a lot of attention among applied scientists and practitioners working in the field of combinatorial optimization and optimization software development since they are encountered in many different manufacturing processes and thus effective solutions to them offer great benefits. In this work, two commonly used heuristic methods for solving production scheduling problems, namely, the Nearest Neighbor (NN) and Ant Colony Optimization (ACO) have been tested on a specific real-life problem and the results discussed. The problem belongs to the class of Asymmetric Travelling Salesman Problems (ATSP), which is known as a hard type problem with no effective solutions for large scale problems available yet. The performances of the Nearest Neighbor algorithm and the Ant Colony Optimization technique were evaluated and compared using two criteria, namely: The minimum value of the objective function achieved and the CPU time it took to find it (including the statistical confidence limits). The conclusions drawn suggest that on one hand the ACO algorithm works better than NN if looking at the achieved minimum values of the objective function. On the other hand, the computational time of the ACO algorithm is slightly longer.

    AB - Production scheduling problems attract a lot of attention among applied scientists and practitioners working in the field of combinatorial optimization and optimization software development since they are encountered in many different manufacturing processes and thus effective solutions to them offer great benefits. In this work, two commonly used heuristic methods for solving production scheduling problems, namely, the Nearest Neighbor (NN) and Ant Colony Optimization (ACO) have been tested on a specific real-life problem and the results discussed. The problem belongs to the class of Asymmetric Travelling Salesman Problems (ATSP), which is known as a hard type problem with no effective solutions for large scale problems available yet. The performances of the Nearest Neighbor algorithm and the Ant Colony Optimization technique were evaluated and compared using two criteria, namely: The minimum value of the objective function achieved and the CPU time it took to find it (including the statistical confidence limits). The conclusions drawn suggest that on one hand the ACO algorithm works better than NN if looking at the achieved minimum values of the objective function. On the other hand, the computational time of the ACO algorithm is slightly longer.

    KW - Ant colony optimization

    KW - Asymmetric travelling salesman problem

    KW - Nearest neighbor

    KW - Production scheduling

    KW - Theory of algorithms

    UR - http://www.scopus.com/inward/record.url?scp=79959807260&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=79959807260&partnerID=8YFLogxK

    M3 - Article

    VL - 40

    SP - 118

    EP - 122

    JO - Information Technology and Control

    JF - Information Technology and Control

    SN - 1392-124X

    IS - 2

    ER -