基于最大熵原理的复合形法及其在边坡稳定分析中的应用

The Complex Method Based on Maximum Entropy Principle and Its Application to the Slope Stability Analysis

将最大熵原理应用于基本复合形法的优化计算中,旨在提高算法的寻优成功率。具体步骤为:选取复形中熵最大的顶点为寻优点;在复形其余各点的加权中心点与寻优点的连线上,搜索比寻优点改善的新顶点;并替换掉寻优点构成新的复形;如此迭代直至复形中的信息被完全利用。采用基本复合形法和基于最大熵原理的复合形法对两个复杂土坡的最小安全系数进行了搜索,结果表明基于最大熵原理的复合形法的全局搜索能力有了明显改善。

Maximum entropy principle is imported into the basic complex method i n order to improve the ratio of successful run. A new search idea is presented a s follows: firstly the search optimum design point is defined as the design poin t whose entropy is the biggest among all the design points of the complex and a new better design point is found on the line through the search optimum design p oint and the integrated centroid of the remaining design points. A new complex i s generated after the search optimum design point is replaced with the new bette r design point. The method iterates until the information of the complex is uti l ized thoroughly thereby proposing a new complex method. The basic complex method and the complex method based on maximum entropy principle are used to locate th e critical failure surface with the minimum factor of safety of two non-homogen e ous soil slopes and the comparisons of the results suggest that the ability of t he complex method based on maximum entropy principle to find global optimum is s tronger than that of the basic complex method.