西南石油大学学报(自然科学版) ›› 1999, Vol. 21 ›› Issue (2): 43-45.DOI: 10.3863/j.issn.1000-2634.1999.02.12

• 地质勘探 • 上一篇    下一篇

综合遗传算法及其在地震波阻抗反演中的应用

尹成1 陈涛1 黄小革1 年永吉1   

  1. 1.西南石油学院勘探系,四川 南充 637001
  • 收稿日期:1999-02-25 修回日期:1900-01-01 出版日期:1999-04-20 发布日期:1999-04-20

HYBRID GENETIC ALGORITHM AND ITS APPLICATION IN SEISMIC WAVE IMPEDANCE IN VERSION

Yin Cheng1 Chen Tao1 Huang Xiao-ge1 Nian Yong-jie1   

  1. Southwest Petroleum Inst
  • Received:1999-02-25 Revised:1900-01-01 Online:1999-04-20 Published:1999-04-20

摘要:

为了有效地克服非线性、多参数、多极值的组合优化问题中遗传算法的早熟收敛现象,引入生物的小生境技术和自然界生物多种杂交方式并存、多个子群体并行以及灾变现象而提出了一种综合的快速基因搜索策略。该算法以多个子群体不同杂交方式并行搜索为主,并结合排序配对的受限交配技术和灾变算子等方法来实现快速的寻优过程。对地震资料波阻抗反演的理论模型试验表明:该算法的收敛能力和计算效率有了明显改善。

关键词: 遗传算法, 早熟收敛, 波阻抗, 反演, 快速搜索遗传算法

Abstract:

A hybrid fast genetic algorithm (H2O) is proposed on the basis of various characteristics of nature evolution. This algorithm is combined with niche technician, a variety of crossover schemes, multi-populations and catastrophism phenomenons. Last sense, it can efficiently eliminate premature convergence of GA on the nonlinear multi-parameter optimization problem. Through testing seism wave impedance inversion, the result is obviously improved compared with the convention genetic algorithm in the convergence ability and computation efficiency.

Key words: algorithm, wave impedance

中图分类号: