模拟退火算法的改进及在静校正中的应用

英文篇名：Improvement of analog annealing algorithm and its application in static correction
中文刊名：大庆石油地质与开发
英文刊名：Petroleum Geology & Oilfield Development in Daqing
作者：李丽 ; 朱国同 ; 陈秀娟 ; 井西利
英文作者：LI Li1 ; ZHU Guo-tong2 ; CHEN Xiu-juan3 ; JING Xi-li1(1.Yanshan University ; Qinhuangdao 066004 ; China ; 2.Exploration and Development Research Institute of Daqing Oilfield Company Ltd. ; Daqing 163712 ; China ; 3.Research Institute of No.1 Geophysical Exploration Company of Daqing ; Daqing 163357 ; China)
中文关键词：模拟退火 ; 静校正 ; 温度 ; 反演
英文关键词：analog annealing ; static correction ; temperature ; inversion
出版日期：2008-10-01
机构：燕山大学;大庆油田有限责任公司勘探开发研究院;大庆钻探集团物探一公司;
年：2008
期：05
出版单位：大庆石油地质与开发

摘要

针对地球物理反演的非线性问题,对模拟退火算法计算效率的影响因素进行了详尽的分析,并通过引入均匀试验设计理论,使在庞大的解空间中随机选取的候选解更加均匀,少量的目标函数计算就能了解目标函数解空间的性质,从而提高模拟退火算法的计算效率。对退火温度的选取方法进行了讨论,并提出了改进方法,使退火温度的选取具有自适应的特点,通过在地震静校正中应用,证明了改进算法的有效性。
In view of nonlinearity characteristic of geophysical inversion,the influencing factors of calculation efficiency in analog annealing algorithm have been analyzed in detail.At the same time,the introduced uniform design and simulation theory makes random chosen candidate solutions more homogeneous in the vast solution space and furthermore a few target function calculations makes clear of the properties of their solution spaces and thus enhances the computation efficiency of analog annealing algorithm.The choosing methods of annealing temperature are also discussed in the paper and moreover the improved method has been presented.All achievements stated above make the choice possess self-adaption characteristic.Through the application in seismic static correction,the improved method has been proven effective.

引文

[1]林依,张中杰,尹成,等.复杂地形条件下静校正的综合寻优[J].地球物理学报,2003,46(1):101-106.
    [2]刘伊克,常旭,王辉,等.三维复杂地形近地表速度估算及地震层析静校正[J].地球物理学报,2001,44(2):272-278.
    [3]井西利.大剩余静校正量求解的两步法[J].石油地球物理勘探,20036,38(1):22-26.
    [4]井西利,杨长春.求解地震静校正问题的双尺度反演方法[J].地球物理学报,2006,49(5):1460-1466.
    [5]师学明,王家映,易远元,等.一种新的地球物理反演方法模拟原子跃迁反演法[J].地球物理学报,2007,50(1):305-312.
    [6]Perez G,Larner K.The quality of the surface consistency assumptionin short wavelength static correction[J].Geophysics,2004,69(4):1091-1101.
    [7]师学明,王家映,张胜业,等.多尺度逐次逼近遗传算法反演大地电磁资料[J].地球物理学报,2000,43(1):122-130.
    [8]Rothman D H.Automatics estimation of large residual statics corrections[J].Geophysics,1986,51:332-346.
    [9]LIUP C,Hartzell S H,Willian Stephenson.Non-linear multiparameterinversion using a hybrid global search algorithm:applications in ref-election seismolgy[J].Geophysics,1995,122:991-1000.
    [10]纪晨,姚振兴.用于地球物理反演的均匀设计优化方法[J].地球物理学报,1996,39(2):233-242.
    [11]王元,方开泰.关于均匀分布与试验设计口[J].科学通报,1981,26:65-70.
    [12]王凌.智能优化算法及其应用[M].北京:清华大学出版社,2001:17.
    [13]艾印双,刘鹏程,郑天愉.自适应全局混合反演[J].中国科学,1998,28(2):1O5-110.
    [14]Larhoven J,Van P I,Aarts E H.Simulated Annealing:Theory andApplication[R].D Riedal Pub.Co.,1986.
    [15]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.