利用协同演化方法实现深基坑支护的分布式优化

基坑支护优化是一个多层次、多参数的复杂系统,常规方法很难体现其层次性.针对深基坑支护问题的层次性和层次性问的相互影响,提出了一种新的协同优化算法:协同演化优化算法.该将现有的协作优化和协同演化结合起来,充分利用了协同演化的层次性和协作优化的并行性.分析了应用于深基坑支护复杂系统设计的协同演化优化算法,讨论了优化子系统间的存在的耦合关系和处理方法,最后以锚固式排桩支护和截水帐幕的并行协同设计,验证了算法的有效性.

New collaborative optimization—co-evolution and its applications in foundation excavation

Foundation excavation optimization design concerns with many combinatorial constraint con-ditions. Its scheme and detail design have hierarchical sequence with discrete design data and extre-mely hard to solve. Co-evolution algorithm combines collaborative optimization with genetic algorithms divides the engineering optimization task to problem space and solution space, they evolve independentlybut interrelatedly, decreases the search time. The principle of co-evolution optimiza-tion and the realization method is introduced, the coupling relationships and the communication technology between sub-optimizers are analyzed. Finally experiment shows its effectiveness in deep foundation excavation.