Study Of A New Optimization Theory Based On GA And Its Application In Arch Gate Optimization

Structural optimization design is a concept relative to traditional structural design. It claims the designer to use optimization theory, digital computer and other moden methods to calculate some schemes in the available region under the design purpose; and select a best theme from them according to the preconcerted goal and request. The practical structural design is a very complicated process. In the design work, one should not only carefully analyze the structure, set up a reasonable, effective and calculatable mathematical optimization model, but also calculate by optimization methods and rationally value and correct the result. Theorially speaking, we always hope to find the best theme in whole region, at least one in local region by optimization design.Genetic algorithm (GA), which simulates the theory of creature evolution, is a bionic algorithms. It has been paid abroad attention in the fields of computer sciense and optimization. It has strong robustness and searching ability in whole region, so it can solve most of the optimization problems in practical projects. Meanwhile, because of the complexity of practical projects, the mathematical model of the structural optimization is always complicated, and its workload is quite heavy. In order to increase computational efficiency and save the storage space in computer, it is always demanded to seek the best theme in a small sample space. But do so, it will restrict the colonial diversity, that is, reduce the searching ability in whole rigion of GA. In order to realize GA in a region with finite samples, this paper brings forward a new GA theory and method, which adopt big mutation operation (BMO) in GA. By this means, we can guarantee the colonial diversity and avoid the phenomenon of the sample prematurity, then improve the GA's efficiency and quality.At the end of this paper, the author designed and realized the new GA in ANSYS environment, and triumphantly used it to the optimization design of an arch gate in water-conservancy project. The perfect result proofs the availability of this algorithms, and finds a new way to optimize huge, complicated structures in practical projects.