An evolutionary computation based on immune operation for constraint optimization problems

Yalong Zhang, H. Ogura, X. Ma, J. Kuroiwa, T. Odaka

    Research output: Contribution to journalArticle

    Abstract

    A large number of infeasible solutions often occur in population of evolutionary Computation (EC) solving the constraint combinatorial optimization problems. The greater the number of infeasible solutions in the population, the worse the performance of ECto search the solution, in the worst case, the algorithm ceases to run. The existing methods, penalty function or multi-objective optimization, can relieve partly the worst case of EC to run. However, they are actually to restrain the infeasible solutions surviving in population, the performance of the EC is not improved. In this study we propose an approach using an important feature of the infeasible solutions in Genetic Algorithms (GA). The approach can not only solve the problem of algorithm ceases to run, but also improve effectively the performance of genetic algorithms searching the optimal solution. From examination of the proposed method on multidimensional knapsack problems, the application of method is effective to solve the problem of algorithm ceases to run as well as to improve clearly the performance of GA.
    Original languageEnglish
    Pages (from-to)404-413
    Number of pages10
    JournalRevista Tecnica de la Facultad de Ingenieria Universidad del Zulia
    Volume39
    Issue number1
    DOIs
    Publication statusPublished - 2016

    Fingerprint Dive into the research topics of 'An evolutionary computation based on immune operation for constraint optimization problems'. Together they form a unique fingerprint.

    Cite this