摘要
基于万有引力搜索算法(GSA)提出了一种改进的万有引力搜索算法(MGSA)。针对GSA在处理优化问题时会出现发散的情况,通过限制粒子的速度同时更改算法中的参数来改善这一问题。算法改进后显著提高了GSA中粒子的探索能力与开发能力,可以获得较强的优化能力。采用MATLAB对8个测试基准函数进行仿真实验,并将该方法引入到边坡稳定分析中。对于边坡稳定性分析,利用MGSA搜索出临界滑动面并结合极限平衡法计算出相应的最小安全系数。结果表明:与GSA法及其他方法相比,MGSA在求解最危险滑动面安全系数时具有更好的优化性能。
Based on the gravitational search algorithm（GSA）, an effective modified gravitational search algorithm（MGSA） is proposed. Since the GSA divergence occurs during optimization process, the new strategy uses a limited velocity to control the particles and changes the parameters in the algorithm to improve this problem. The modified algorithm improves the exploration and exploitation ability significantly, and has better optimization ability. The performances of the MGSA and GSA are testified by MATLAB on a suite of eight benchmark functions, and the results are compared. Then the method is used to analyze slope stability. For this problem, the MGSA is used to search the critical slip surface and to calculate the corresponding minimum safety factor based on the limit equilibrium method. The simulated results illustrate that the MGSA has better optimization performance in searching the critical slip surface and calculating the safety factor compared with the GSA and other methods.
出处
《岩土工程学报》
EI
CAS
CSCD
北大核心
2016年第3期419-425,共7页
Chinese Journal of Geotechnical Engineering
关键词
万有引力搜索算法
边坡稳定
临界滑动面
安全系数
基准函数
gravitational search algorithm
slope stability
critical slip surface
safety factor
benchmark function
