论文部分内容阅读
提出了边坡可靠度分析的一种新的全局优化方法——认知聚类分区方法。该方法主要包括5个步骤:分区、随机抽样、计算极径L、回代、计算可靠指标及验算点。给出了相应的计算流程图,并编写了基于C语言的计算程序KCPREL。最后,以岩质边坡稳定可靠度问题为例证明了所提方法的有效性。结果表明,认知聚类分区方法能同时计算出可靠指标和验算点,并能获得全局最优解。该方法的计算精度和蒙特卡洛模拟方法相当,计算效率远远高于传统的蒙特卡洛模拟方法。此外,该方法在分析含有复杂的隐式及非线性功能函数的边坡稳定可靠度问题方面体现出明显的优越性。等步长认知聚类分区方法能全面且均匀地搜索角度,从而得到更准确的验算点。为了保证足够的计算精度及减小计算量,建议步长取10o以内。
A new global optimization method for slope reliability analysis is proposed, which is called cognitive cluster partitioning method. The method mainly includes five steps: zoning, random sampling, calculating the diameter L, back to generation, calculating reliable indicators and checking points. The corresponding calculation flow chart is given and a C program based on C language KCPREL is written. Finally, the reliability of the rock slope is taken as an example to demonstrate the effectiveness of the proposed method. The results show that the cognitive clustering method can calculate reliable indices and checking points simultaneously and get the global optimal solution. The accuracy of this method is equivalent to that of Monte Carlo simulation, and the computational efficiency is much higher than the traditional Monte Carlo simulation method. In addition, this method shows obvious superiority in analyzing the stability of slope with complex implicit and nonlinear function. The equal-step cognition clustering method can search angles comprehensively and uniformly, so as to obtain more accurate checking points. In order to ensure sufficient calculation accuracy and reduce the amount of calculation, it is recommended to take steps within 10o.