Summary: | 碩士 === 國立中山大學 === 機械工程學系 === 87 === The purpose of this study is to develop a paralleled and multilevel technique in the traditional genetic algorithms in order to establish a faster and more efficient new method, distributed hybrid multilevel genetic algorithm.
Distributed genetic algorithms can slove the problem of traditional sequential genetic algorithms, such as premature convergence, large number of function evaluations, and a difficulty in sentting parameters. By using several concurrent sub-population, distributed genetic algorithms can avoid premature convergence resulting from the single genetic searching environment of sequential genetic algorithms. It is useful to speed up the operation rate of joining timely multilevel optimization with paralleled genetic algorithms. Because multilevel optimization can resolve one problem into several smaller subproblems, each subproblem is independent and not interference with one another. Then the subsystem of each level can be connected by sensitivity analysis. So we can solve the entire problem. Because each subproblem contains less variables and constrains, it can achieve the faster converge rate of the entire optimization. Therefore, paralleld hybrid genetic multilevel algorithm consists of both advantages of the distributed genetic algorithms and the multilevel optimization methods.
This method can be applied to several design optimization problems of trusses because of the reduction of execution time of paralleled computing. In addition, the efficient searching function and results are also identified in this study. In short, the paralleled hybrid multilevel genetic algorithms are efficient tools in solving structural optimization problems.
|