Applications Of Genetic Algorithm In Nonlinear Optimization Of Geotechnical Engineering

There are many geotechnical engineering problems can be summed up as optimization, prediction, model building and other issues, which solved by traditional system engineering method sometimes may have certain limitations. Therefore, introducinng genetic algorithm to geotechnical engineering disciplines and doing interdisciplinary research is very necessary, whatever from the sake of the development law of the disciplines, or from the actual needs of disciplines emerging. Therefore, for trying to provide a new method, the application of genetic algorithm in geotechnical engineering problems which need to be optimizd or to be resolved was promoted in this paper. Specifically, the major study propersed in this paper mainly including the followings: (1) Basd on the traditional design methods, using genetic algorithm to optimize the structure parameters of the cantilever pile model, in which have many complex nonlinear constraints. (2) On the analysis of prediction problem of geotechnical engineering, three prediction model which based on genetic algorithm was proposed, including the exponential curve model of ultimate bearing capacity of single pile, the Poisson curve model of foundation settlement and the Verhulst model of foundation settlement. (3) The van genuchten equation (VGE) of which is used to describe the feature of soil water movement is super-set and nonlinear. Using traditional method calculating the VGE's parameters often results with a high margin of error because of complication and artificial factors.Therefor, the method which using genetic algorithm to estimate parameters of van genuchten equation was proposed. (4) There is no direct hydraulic calculation formula for critical water depth of arched tunnel. In order to substitute for the traditional inefficient method such as trial-and-error method, graphical method .etc, the non-linear optimization model was propsosed throuth mathematical transform and the solution - real coding based accelerating genetic algorithm was offered. By analysis, the results show that genetic algorithm is an effective algorithm in solving the optimization problem of geotechnical engineering with many good properties such as high efficiency, high precision, fast computing speed and applicable to the complex optimization problems. Therefore, genetic algorithm have a wide range application value and prospects in geotechnical engineering disciplines, it is worth to be premoted for further applications.