摘要
在两层协同演化模型中,细部演化是影响整个系统计算效率的关键因素,为此在两层协同演化的基础上,进一步提出了分解协同演化模型,将基坑支护细部优化设计系统分解成几个低维、与各子工程对应的、相对独立的子优化器和一个进行全局协调优化的低维系统优化器.并建立分布式并行计算环境,设计基于遗传算法的基坑支护分布式三层协同演化系统(DODRP),实现了三层协同演化模型的分布式计算.运用DODRP系统,对某国际贸易大厦深基坑支护工程进行优化设计,验证了系统的优化设计功能.
Detail evolution is the key factor that influences the computing efficiency of the whole system in the two level co-evolution model.The decompose-collaborated evolution model based on the two-level co-evolution model is advanced.This model decomposes the detail optimization design of the retaining and protecting system into several low dimensional and independent sub-optimizers that represent the corresponding sub-projects, and a low dimensional system optimizer for global coordinated optimization. A distributed parallel computing environment is constructed.Based on genetic algorithm'the Distributed three-level co-evolution system for the Optimization Design of the Retaining and Protecting system for foundation pit'(DODRP) is designed and the distributed computing of three-level co-evolution model is implemented.The DODRP system has been successfully used in the optimization design of the project of 'the retaining and protecting engineering for deep foundation excavation of an International Trading Building'.It is proved that the system performance is good in the optimization design.
出处
《桂林工学院学报》
CAS
2004年第4期435-440,共6页
Journal of Guilin University of Technology
基金
国家自然科学基金资助项目(59868001)
关键词
基坑
支护
优化设计
遗传算法
分布式计算
foundation pit
retaining and protecting
optimization design
genetic algorithm
distributed computing