GLOBALIOJO OPTIMIZAVIMO UŽDAVINIŲ SPRENDIMAS NAUDOJANT GENETINIUS ALGORITMUS
GENETIC ALGORITHMS FOR SOLVING GLOBAL OPTIMIZATION PROBLEMS

Summary/Abstract: The authors examine the theoretical aspects of solving global optimization problems and analyse how global optimisation can be used in bored pile foundations. The position of bored pile foundations was determined with FORTRAN. When the C++ code was optimized the overall performance of the program decreased only by 0,008s. Optimization problems were solved with genetic algorithms, the time taken to execute optimization and genetic algorithms was compared. It was found that genetic algorithms have no impact on computing resources. Eight strategies for using various combinations of genetic algorithms were tested in order to identify the most effective one, the findings were compared with the findings of other scientists. The result when a global optimization problem was solved with the proposed genetic algorithm was by 1.9% better than that using the Bayesian method but by 4.6% worse than using a simulated annealing method described in literature as the best one.

• Details
• Contents