改进遗传算法在房建深基坑支护结构设计中的应用

Application of improved genetic algorithm in the design of retaining structure for deep foundation pit in residential buildings

住宅楼工程施工中经常采用复杂双排灌注桩且一桩两用的方式设计基坑支护结构,但工程实践中基坑施工具有明显的不确定性,各方面因素共同作用下导致数据模拟数值和理想数值之间通常存在鲜明差异。为此,以张家口宇鑫小区二期项目为例,研究基于改进遗传算法在深基坑支护结构设计中的应用,通过构建模型时取桩体嵌入地面的深度和桩体弯矩作为约束条件并设置惩罚函数,明确模型建立中的约束问题,同时结合施工现场基坑断面处理图形结构,通过在基坑中设置钢板桩支护结构以约束基坑变形等问题,并采用高压旋喷桩等方式确保基坑支护结构的防渗防水效果。实验结果显示,改进遗传算法的应用能够有效提高地表沉降、桩体的水平位移和地表沉降与理想数值之间的契合性,大幅度提高数值模拟可信度。

Complex double-row cast-in-place piles are often used in the design of retaining structure for deep foundation pit in the construction of residential buildings,however,there are obvious uncertainties in the construction of foundation pit in engineering practice,and there are obvious differences between the numerical simulation and the ideal numerical simulation.This paper studies the application of improved genetic algorithm in the design of deep foundation pit bracing structure of the second phase project of Yuxin district in Zhangjiakou.Taking the depth of the pile embedded in the ground and the bending moment of the pile as the constraint conditions and setting the penalty function,the constraint problems in the model establishment are clarified,in order to control the deformation of the foundation pit,the steel sheet pile support structure is set up in the foundation pit,and the high-pressure jet grouting pile is used to ensure the anti-seepage and water-proof effect of the foundation pit support structure.The experimental results show that the application of the improved genetic algorithm can effectively improve the surface settlement,pile horizontal displacement and the agreement between the surface settlement and the ideal value,and greatly improve the reliability of the numerical simulation.