首页 | 本学科首页   官方微博 | 高级检索  
     

基于复合形算子的基础支护桩优化设计智能算法研究
引用本文:钟登华,梅传书,韩圣章.基于复合形算子的基础支护桩优化设计智能算法研究[J].工程力学,2003,20(1):80-85.
作者姓名:钟登华  梅传书  韩圣章
作者单位:天津大学建工学院,天津,300072
基金项目:国家自然科学基金(50179023),国家教育部《高等学校骨干教师资助计划》,《高等学校优秀青年教师教学科研奖励计划》项目
摘    要:本文通过遗传算法和传统复合形搜索法相结合,基于对遗传算法算子计算结构的调整,并将遗传算法与神经网络相结合,提出并研究了一种新的优化设计方法,协同求解复杂工程中的优化问题。并针对悬臂式支护桩的优化设计的数学模型,采用该算法进行了优化设计分析;计算结果表明,该算法可克服遗传算法最终进化至最优解较慢和人工神经网络易陷入局部解的缺陷,具有较好的全局性和收敛速度。

关 键 词:复合形算子  遗传算法  人工神经网络  悬臂桩  优化设计
文章编号:1000-4750(2003)01-080-06

AN INTELLIGENT ALGORITHM BASED ON COMPLEX SHAPE OPERATOR FOR OPTIMUM DESIGN OF SUPPORTING PILES
ZHONG Deng-hua,MEI Chuan-shu,HAN Sheng-zhang.AN INTELLIGENT ALGORITHM BASED ON COMPLEX SHAPE OPERATOR FOR OPTIMUM DESIGN OF SUPPORTING PILES[J].Engineering Mechanics,2003,20(1):80-85.
Authors:ZHONG Deng-hua  MEI Chuan-shu  HAN Sheng-zhang
Abstract:An algorithm for optimum design is developed through combining genetic algorithm with traditional complex shape search method. Neural networks are incorporated into the genetic algorithms and the computational procedures of genetic algorithms operators are adjusted. A practical deep excavation case is studied using the present algorithm. It is shown that the defects of genetic algorithms and artificial neural networks such as slow convergence and immersion into local solutions are done away with. Good global performance and satisfactory convergence rate are achieved.
Keywords:complex shape operator  genetic algorithm  artificial neural network  cantilever supporting piles  optimum design
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号