摘要
岩土边坡的稳定性取决于边坡的几何形状和岩土材料的抗剪指标参数。由于尺寸效应,根据实验室小尺寸试件获得的参数具有某些局限性,反分析方法为解决这一课题提供了新的途径。根据实验观测岩土边坡滑移面的形状参数,建立了应用混合遗传算法识别岩土材料的抗剪指标参数的方法。遗传算法的特点之一是不需要计算目标函数的导数。为了提高参数识别方法的有效性、收敛速度和鲁棒性,将梯度搜索方法和遗传算法有机地结合在一起,使得所建立的反演策略具有梯度搜索方法和遗传算法的各自特性。研究了观测信息的不确定性对参数反演结果的影响。通过一个数值算例对比,显示了所建立参数识别方法的实用性和有效性。
Rock slope stability depends on the strength features of the rocks. Determination of the strength of rock masses is difficult since the size of representative specimens is too large for laboratory testing. This difficulty can be overcome by using the parameter identification procedures. The inverse problem of geotechnical parameters identification is treated as the optimization problem. The hybrid genetic algorithm is presented and used for the identification of the shear strength parameters of geotechnical materials according to the observed shape of slip surface of rock slope. Genetic algorithm (GA) is a powerful alternative to traditional optimization methods which are too restrictive. One of the main advantages of GA is that it requires no information about the objective function. In order to identify parameters efficiently and in a robust manner, the gradient search method and genetic algorithm are combined to produce an adaptive procedure that has the merits of the two methods. The validity and efficiency of the proposed procedure are shown by an example involving parameter identification. The results obtained are compared with earlier results obtained by other methods.
出处
《岩石力学与工程学报》
EI
CAS
CSCD
北大核心
2005年第4期676-680,共5页
Chinese Journal of Rock Mechanics and Engineering
基金
国家自然科学基金资助项目(10072014)
关键词
岩土力学
参数识别
岩土材料
混合遗传算法
观测误差
Genetic algorithms
Gradient methods
Inverse problems
Parameter estimation
Shear strength
Slope stability
Soil mechanics
