Study On Optimization Design Of Supporting Structure For Deep Excavation Based On Improving Genetic Algorithm

Deep excavation supporting structure is a provisional establishment. Its function is to protect the stability of deep excavation and the safety of the structures and establishments around the deep excavation. It will be wasteful if the selection and design of the supporting structure is too safe, so optimization design is very important when we design the supporting structure of deep excavation.Currently, the types of supporting structures are so many that it will still be a difficult problem of how to choice a appropriate supporting structure and gain the accurate parameters. This article analyzes several cross-sectional supporting structures, and discusses their some parameters. This article investigates several destructive forms of supporting structures, and expatiates some ways of optimization design which we use frequently now, and shows the advantages of genetic algorithm. This article also sets up a mathematic model of deep excavation retaining system, and uses GA to design a supporting structure. The result shows that using GA to design supporting structure is doable.How to deal with the restrictions of deep excavation is a very difficult problem. This paper analyzes the restrictions of pile-support structure detailedly, uses a castigatory function to deal with the restrictions, and adds the castigation to the function. Now the problem has changed to solve a minimal problem, and GA can be used to the optimal design successfully.SGA has many disadvantages such as premature and poor search ability. This paper ameliorates the disadvantages . Now depict the betterments :1. Adopting the strategy of holding the best chromosome to make sure the best chromosome come into the next circulation;2. Using the adaptive idea to ameliorate the crossover and recombination by different fitness. This can hold the multiformity of population and debase the possibility of destroying the better fitness chromosome;3. The broadly search ability of GA is better than the local search ability, so this paper combine hill-climbing and SGA to form a new hybrid genetic algorithm. When the evolution achieves the definite number or prescriptive numerical value, we beginthe evolution achieves the definite number or prescriptive numerical value, we begin to search the better solution by hill-climbing. This can add the local search ability and make the solution more accurate.This paper develops a GA software system of deep excavation by Visual Basic. Using the mathematic model we have established, this paper designs a supporting structure, and gains the different results of SGA and the adaptive hybrid genetic algorithm. From the results we can see that the adaptive hybrid genetic algorithm is better than SGA. This shows that the idea of this paper is right.