多目标约束下的土石坝抗液化措施数值建模与优化分析方法

Numerical modeling and optimization analysis of anti-liquefaction measures of earth rock dam under multi-objective constraints

土石坝坝基地震液化问题严重威胁工程安全。现阶段的液化研究主要集中于液化机理分析,而在抗液化措施的设计优化方面,现有土石坝工程主要依靠以往的工程经验来处理坝址的液化土层。本文构建了以土石坝坝体坡面水平位移梯度、坝顶震陷和抗液化措施成本最小化为优化目标的数学模型,提出了以静动力抗滑稳定安全系数、地基最大液化范围、液化深度、典型点地震过程中平均液化度和震后永久水平位移为约束条件的土石坝压重平台和坝基置换参数优化设计方法;采用完全非线性动力算法和多目标遗传算法,通过VC++集成FLAC3D进行开发,实现了综合考虑土石坝抗液化措施的安全性能与经济效益的多目标优化设计。该方法应用于工程实例中,采用存档微遗传算法对压重平台及地基开挖置换的9个控制性设计参数进行优化。结果表明,优化后的方案相较初始方案使上下游坝体坡面水平位移梯度指标降低了16.59%和24.33%,坝顶震陷减少了21.03%,工程造价节约了18.53%。所提出的模型与方法为改善土石坝坝基液化问题提供了新的思路。

The seismic liquefaction of the foundation of earth rock dam is a serious threat to the safety of projects.The current research of liquefaction mainly focuses on the analysis of liquefaction mechanism,while in the design and optimization of anti-liquefaction measures,the existing earth rock dam projects mainly rely on the previous experience to deal with the liquefiable foundation of the dam site.A mathematical model is built to minimize the horizontal displacement gradient of earth rock dam slope,seismic subsidence on the top of the dam and the cost of anti-liquefaction measures.An optimal design method of parameters for ballast platform and foundation replacement of earth rock dam is proposed,which is constrained by the safety factor of static and dynamic anti-sliding stability,the maximum liquefaction range of foundation,the depth of liquefaction,the average degree of liquefaction of typical points in the process of earthquake and the permanent horizontal displacement after earthquake.The fully nonlinear dynamic algorithm and multi-objective genetic algorithm are used and FLAC3D is integrated by VC++,which realize the multi-objective optimization design considering the safety performance and economic benefits of anti-liquefaction measures of earth rock dam.This method is applied to a practical project,and nine controlling design parameters of the platform and the foundation excavation replacement are optimized by using the Archivebased Micro-Genetic Algorithm(AMGA).The results show that compared with the original scheme,the horizontal displacement gradient of upstream and downstream dam slope decreased by 16.59%and 24.33%,the seismic settlement of the dam crest is reduced by 21.03%,and the project cost reduced by 18.53%.The proposed model and method provide a new way to reduce the liquefaction of earth rock dam foundation.