A Multiuser Manufacturing Resource Service Composition Method Based on the Bees Algorithm

Article English OPEN
Yongquan Xie ; Zude Zhou ; Duc Truong Pham ; Wenjun Xu ; Chunqian Ji (2015)
  • Publisher: Hindawi Limited
  • Journal: Computational Intelligence and Neuroscience, volume 2,015 (issn: 1687-5265, eissn: 1687-5273)
  • Related identifiers: pmc: PMC4538778, doi: 10.1155/2015/780352
  • Subject: R858-859.7 | Research Article | Computer applications to medicine. Medical informatics | Neurosciences. Biological psychiatry. Neuropsychiatry | RC321-571 | QA76 | Article Subject

In order to realize an optimal resource service allocation in current open and service-oriented manufacturing model, multiuser resource service composition (RSC) is modeled as a combinational and constrained multiobjective problem. The model takes into account both subjective and objective quality of service (QoS) properties as representatives to evaluate a solution. The QoS properties aggregation and evaluation techniques are based on existing researches. The basic Bees Algorithm is tailored for finding a near optimal solution to the model, since the basic version is only proposed to find a desired solution in continuous domain and thus not suitable for solving the problem modeled in our study. Particular rules are designed for handling the constraints and finding Pareto optimality. In addition, the established model introduces a trusted service set to each user so that the algorithm could start by searching in the neighbor of more reliable service chains (known as seeds) than those randomly generated. The advantages of these techniques are validated by experiments in terms of success rate, searching speed, ability of avoiding ingenuity, and so forth. The results demonstrate the effectiveness of the proposed method in handling multiuser RSC problems.
  • References (22)
    22 references, page 1 of 3

    ( 60 d

    n 40 ie

    s 20 d

    F 0 100 90 )(%80 e tra70 s sce60 c Su50 40 30 80

    v 40 icen

    s 20 adu

    F 0 100 90

    ( 60 d

    n 40 ie

    s 20 d

  • Metrics
    No metrics available
Share - Bookmark