| 地球物理非线性反演模拟退火法的改进 |
| |
| 引用本文： | 姚姚.地球物理非线性反演模拟退火法的改进[J].地球物理学报,1995,38(5):643-650. |
| |
| 作者姓名： | 姚姚 |
| |
| 作者单位： | 中国地质大学, 武汉 430070 |
| |
| 摘 要： | 在地球物理非线性反演方法中，模拟退火法是一种较先进的启发式蒙特卡洛（ＭｏｎｔｅＣａｒｌｏ）方法．但是，在处理实际资料时，该方法存在着计算效率不够高的缺点，有时还会失效．为此，从模拟退少法的关键问题──最低温度的选择入手，根据模拟退火法与统计力学的吉布斯－马尔柯夫（Ｇｉｂｂｓ－Ｍａｒｋｏｖ）模型之间的关系导出临界温度的近似表达式；用此式分析目标函数超曲面形状对模拟退火法计算的影响；提出利用模糊先验信息确定最低温度、改造目标函数等改进措施．
|
| 关 键 词： | 临界温度 模糊目标函数 模拟退火法 非线性反演 |
| 收稿时间： | 1994-07-25 |
IMPROVEMENT ON NONLINEAR GEOPHYSICAL INVERSION SIMULATED ANNEALING |
| |
| YAO YAO.IMPROVEMENT ON NONLINEAR GEOPHYSICAL INVERSION SIMULATED ANNEALING[J].Chinese Journal of Geophysics,1995,38(5):643-650. |
| |
| Authors: | YAO YAO |
| |
| Institution: | China University of Geoscience, Wuhan 430070 |
| |
| Abstract: | Among nonlinear geophysical inversion methods,simulated annealing is a modern heuristic Monte Carlo approach.However,its calculation efficiency is not high enough and sometimes it is invalid when processing real data.Therefore we begin with selection of minimum temperature──the key of simulated annealing,derive the formula of approximately calculating critical temperature from the relationship between simulated annealing and Gibbs-Markov model in statistical mechanics,analyse the influence of objective function super-surface shape on the performance of simulated annealing using the formula,and propose the improvement on simulated annealing by determing the minimum temperature and reforming objective function using prior fuzzy information. |
| |
| Keywords: | Critical temperature Fuzzy objective function Simulated annealing Nonlinear inversion |
| 本文献已被 CNKI 维普 等数据库收录！ |
| 点击此处可从《地球物理学报》浏览原始摘要信息 |
| 点击此处可从《地球物理学报》下载免费的PDF全文 |
