边坡稳定可靠度分析的神经网络法_谭晓慧.pdfVIP

第18 卷第3 期 安徽建筑工业学院学报( 自然科学版) Vol. 18 No. 3 2010 年6 月 Journal of Anhui Inst itute of Architecture Industry Jun. 2010 谭晓慧, 李 丹, 吕晓光 ( , 230009) : , GEO SLOPE , ,RBF (RBF MCSM) , , GEO SLOPE GEO SLOPE Slope/ w , ( RBF MCSM ) , , ( ) , RBF M CSM , RBF MCSM : ; ; ; GEO SLOPE :TU43 :A : 1006 4540( 2010) 03 072 05 Reliability analysis of slope stability based on neural network TAN Xiao hui, LI Dan, LV Xiao guang ( College of Recourses and Environment Engineering, Hefei University of Technology, Anhui Hefei 230009, China) Abstract: The conventional reliability analysis methods of slope stability, which can be used to analysis the slope stability, is very cumbersome because it must interact the main program and the process of solving the safety factor with the GEO SLOPE software. To improve this situation, the method that RBF neural network based on Monte Carlo simulation method ( RBF MCSM) has been applied in this paper. You should prepare all samples with the GEO SLOPE software before you run the reliability a nalysis program. In this paper, firstly, the Slope/ w module in the GEO SLOPE software had been used to solve the safety factor of four groups in Piyuan Slope before and after supporting. After that, the author used the method of Monte Carlo simulation based on radical basis function neural network to fit a performance function, and solved the probability of slope failure and reliable indicator. Finally the author compared the results with the conventional analy