Study On Parameter Inversion Of Underground Engineering Based On Improving Genetic Algorithm

The environment of the underground engineering is sophisticated and changeable, every parameter related to the Rock and Soil is defective because the work quantity is big and the calculation is complex. Though the calculation theory is gradually perfect and applicates widely in the engineering, there is still some questions. The main reason is that the distributation scale and quantity of the previous grand stress imported as the given factors are usually not definite, meanwhile the different quantity of the strate parameter such as E and n are radom to the different zone. However, we usually simulates the conditions of the engineering working geology which let to the destruction after the warp caused by the strength with ideal model,and the destruction conditions the warp of the strate caused by the strength is short of the judgement principle, etc. So, the result usually can't accord with the truth.This thesis contrasts and analyses the different calculation-method, and improves the Genetic Algorithm in the parameter inversion of the Rock and Soil engineering. The strategies are mainly in the following three aspects:1 , new searching way is Compositely searching Genetic Algorithm which is made up of the acceleration method and Genetic Algorithm when it partially calculates in the later time;2, index function is inducted in gengetic operators, at the same time repeated select and double exchange pool are used;3, Splicing crossover, which reserves some new evolution factions, improves authority function accaunting the capabilities and kinds.The Improving Genetic Algorithm, carrying on FORTRAN language, is used in SuiBuya water conservancy hub. Based on the ANSYS language procedure calculation , the stress parameter vibration of the underground carvera accords with the truth and the answer is correct.