Inverse Analysis of Slope Engineering Based on Inverse Reliability
|School||Central South University|
|Course||Underground Space Science and Engineering|
|Keywords||inverse analysis inverse reliability response surface method genetic algorithm inverse problem|
Most reliability problems in slope engineering are positive reliability problems. It means we are to compute the reliability of slope on the condition parameters are given, while many problems arise which need to compute some parameters to satisfy specified reliability index. For example, the reliability index of slope of strip mine is changing in the excavation, and roads with different levels acquire different reliability index so their design parameters are different. Problems like these are called inverse reliability problems. Based on the needs for parameter inversion, this paper refer to the application of inverse problem in other subjects, especially in structural mechanics, and raise the inverse reliability method for slope engineering when inverse model established and solving process derived.In this article slope’s minimum safety factor is computed by simplified Bishop method, and the parameters and effect of boundary condition’s change on result is involved in this program, so it can give a more precise evaluation of stability problem. Response surface method is applied here to establish explicit slope stability limit state equation for that the limit state equation obtained by Bishop method is implicit and cannot be used directly. Due to the development of intelligent optimization technique, especially the genetic algorithms’wide application for its excellent performance and significant effects, so genetic algorithm is applied in this paper, and good results are achieved.Several examples are computed and results compared with each other. It is approved that the results is close to the actual so the accuracy requirement is satisfied. Above all, this article provides a thinking way, also a platform for development of inverse problems in slope engineering.