Optimal Design Of Cluster Pile-Steel Tube Supporting Structures Based On Genetic Algorithm

Nowadays, the types of supporting structures in the foundation pit are so many that it will still be a difficult problem of how to choice a appropriate supporting structure and gain the accurate parameters. Cluster pile/steel tube supporting structures are used widely in foundation pit engineering owing to less influence on environments and more adjustability to the stratum.At first, the category, breaking forms, function and characteristics of the supporting structures are introduced, and a series of problems are analyzed. The methods for the earth pressure calculation and for the design of the supporting structures in deep foundation pit are analyzed in detail, and the difference is compared between the traditional methods and the modern ones, it safely concludes that the modern methods overmatch the traditional ones. Particularly discussed the theory and implementation of the genetic algorithms. Generic Algorithm is a global algorithm which uses natural law" survival of the fittest, extinction of the unfitness" and natural selection mechanism to guide and determinate search direction. Comparing with other optimization methods, it needs few mathematical requirements, so it appropriates to optimization design of supporting structures for foundation pit. The question on how to choose the crossover probability, the mutation probability, the scale of population and the numbers of the generation are discussed. Then, the mathematics model of the cluster pile/steel tube supporting structures based on genetic algorithm for optimal design is established. After integrating the calculation method for increment considering the behavior under staged excavation and determining the decision variables, objective function, constraint conditions, and the computer programming , the optimization about cluster pile/steel tube supporting structures is successfully carried out.In order to expedite the speed of convergence, reserving the best individual do not take part in the inheritance of the next generation, thus increasing the speed of convergence and enhancing the efficiency to cut down the running time. It is proved by the practical engineering that the modified genetic algorithm has better convergence performance.