论文部分内容阅读
将模糊随机理论与改进的响应面模型结合求解滑坡的可靠性指标。首先,考虑到滑带土力学参数的模糊随机性,非线性程度较高,采用正态模糊隶属函数对参数样本进行模糊-随机处理,并作为优化模型的约束函数;然后,采用过无差点的改进响应面法将功能函数显式化为二阶近似函数,黏聚力和内摩擦角约束函数显式化为线性函数,基于二次规则方法,采用理性运动极限计算方法,寻找合理、有效而稳定的优化策略求解滑坡的目标功能函数;最后,采用一次二阶矩法计算滑坡的可靠性指标。改进的响应面法在迭代过程中通过中心展开点,优化过程可以收敛到精确的局部最优点,同时理性运动把每次寻优的空间缩小,在建立近似显函数展开点或构造点的邻域进行寻优,避免了优化迭代振荡、收敛速度缓慢或不收敛的缺陷,提升了优化速度并保证求解的有效性和准确性。计算方法保证了改进响应面法的精度,求得的可靠性指标考虑了参数模糊不确定性的影响,研究成果为响应法的改进并应用于滑坡工程的可靠性评价领域起到一定的推动作用。
Combining fuzzy stochastic theory and improved response surface model to solve landslide reliability index. First of all, considering the fuzzy randomness and high degree of nonlinearity of the mechanical parameters of sliding zone, the normal fuzzy membership function is used to carry out the fuzzy-random process on the parameter samples as the constraint function of the optimization model. Then, The response surface method is adopted to make the function function explicit as a second order approximation function. The cohesion force and internal friction angle constraint function are explicitly linearized. Based on the quadratic rule method, the rational motion limit method is used to find the reasonable and effective Stable optimization strategy is used to solve the target function of landslide. Finally, a second-order moment method is used to calculate the reliability index of landslide. In the iterative process, the improved response surface method expands the point through the center, and the optimization process can converge to the exact local optimal point. Meanwhile, the rational motion reduces the space of each optimization, and establishes the approximate function expansion point or the neighborhood of the construction point It can avoid the problem of optimizing iterative oscillation, slow or non-convergent convergence, improve the optimization speed and ensure the validity and accuracy of the solution. The calculation method ensures the accuracy of the improved response surface method and the reliability index obtained considers the influence of the fuzzy uncertainty of the parameters. The research results provide a certain impetus to the improvement of the response method and its application in the reliability evaluation of landslide engineering .