【摘要】 针对成都地区土层特有的沉积环境及物理力学性质的差异性,采用Martin-Davidenkov模型和阻尼比经验公式,使用基于实码、并引入了更新概率(Pu)的加速遗传算法(AGA),编制了相应的C++程序,反演计算出成都地区各类沉积土的动剪切模量比G/Gmax和阻尼比λ与剪应变幅值γ的关系曲线模型的参数值。结果表明,RAGA收敛速度快,计算精度高,是一种既可以较大概率搜索全局最优解,又能进行局部细致搜索的非线性优化方法。更多还原

【Abstract】 In light of unique sedimentary environment of Chengdu region soil and the differences of the physical and mechanical properties,Martin-Davidenkov model and the empirical formula of damping ratio are used,the corresponding C++ program is done through fact-based code and the introduction of the updated probability of accelerating genetic algorithm,by inversion,we can calculate dynamic shear modulus ratio G/Gmax and damping ratio λ and the shear strain amplitude γ of the curve model parameters of the various types of deposits in Chengdu.The results show that:because of its fast convergence and high accuracy,RAGA is a nonlinear optimization method that not only can search the global optimal solution in greater probability,but also can search local details.更多还原