基于MPGA的边坡三维稳定性分析

Three-dimensional analysis of slope stability based on MPGA

目前三维极限平衡法在边坡稳定分析中处于主导地位,为了达到静定可解,各种三维极限平衡方法均引入大量的假定,削弱了该方法的理论基础的应用范围.为了解决以上问题,将多种群遗传算法(MPGA)引入三维边坡的稳定性分析,结合有限元计算建立一种基于MPGA边坡三维稳定性分析的新方法.通过边坡的数值应力场求解安全系数,并为多种群遗传算法构建适应度函数,再利用多种群遗传算法为安全系数的计算提供滑移面.根据滑移面的发展趋势动态产生滑移面,增加滑移面的约束条件以保证分析的高效性.最后通过三维边坡经典算例分析验证该方法以及程序的合理性.该方法摈弃了三维极限平衡法所做的种种假定,仅需要1次有限元计算结果,而且计算结果可靠且易于编程.

At present,the 3-D limiting equilibrium method is in a dominant position in the stability analysis of slope.However,various 3-D limiting equilibrium methods introduce a mass of supposition into themselves to achieve a statically definite solution,so that the application scope of the theoretical basis of this method is weakened.To solve aforementioned problem,a multi-population genetic algorithm（MPGA）is introduced into three-dimensional slope stability analysis and an MPGA-based new method for 3-D analysis of slope stability is established with finite element method.The safety factor of slope is found with numerical stress field of slope and the adaptability function is constructed for MPGA.The MPGA is also used to provide the slip surface for the safety factor calculation.To ensure high efficiency to analysis,the initial slip surface is generated dynamically according to its development trend and a constraint condition is also added to the slip surface.Finally,by means of classic case of 3-D slop stability analysis,the rationality of aforementioned method and programs is analyzed and verified.In this method,various assumptions for the method of three-dimensional limiting equilibrium are abandoned,and only a single finite element calculation is needed to get the calculation results with reliability and easy programming.