Horizontal Displacement Analysis And Prediction Of Composite Soil-nail Wall

The application of soil-nail wall in our country is having a rapid develpment, lots of engineering experiences and research results has been presented. However, researches and analyses on displacement of soil-nail wall are comparely not enough. This situation doesn't meet the need of more and more strict displacement control in excavation engineerings. To against this, this paper analyses the horizontal displacement of soil-nail wall by introducing the artificial intelligence idea, and also creats a method to compute the horizontal displacement of composite soil-nail wall according to the displacement characteristic of composite soil-nail wall.Firstly, the author represents the characteristic of excavation engineerings, styles, computing theories and applicability of frequently used retaining walls are also illuminated.Then introduces the basic theories of BP artificial neural-net work, its applications and present research state in geotechnic engineerings, modifies currently used modeling methods, so as to easily take into account of the influence of desigh, construction and soil-layers parameters. Under Matlab6.5, use this method to predict horizontal displacement of soil-nail wall. By engineering example, this method is proved to be reasonable and comparely better than traditional ones.Because the displacement of composite soil-nail wall is similar to the normal internal- stayed retaining wall, elastic-foundation beam method is introduced into the method of computing composite soil-nail wall's displacement, simplified soil-pressure distribution model and soil-nail's staying-stiffness computing method which are associated with the characteristic of composite soil-nail wall are also presented, then a method of computing composite soil-nail wall's horizontal displacement is created.by back-analysis use BP neural-net work, the change regularity of soil-nail's staying-stiffness in this method is obtained. Finally the author proves the result and indicates that this method has a good precision and applicability by several engineering projects.