基于凸集比例因子和WOA-Kriging模型的重力坝非概率可靠性分析  

Non-probabilistic Reliability Analysis of Gravity Dam Based on Convex Set Scale Factor and the WOA-Kriging Model

作　　者：刘要来 王堡生 周红波 赵二峰[2] 李章寅 LIU Yao-lai;WANG Bao-sheng;ZHOU Hong-bo;ZHAO Er-feng;LI Zhang-yin(Power China Zhongnan Engineering Corporation Limited,Changsha 410014,China;College of WaterConservancy and Hydropower Engineering,Hohai University,Nanjing 210098,China)

机构地区：[1]中国电建集团中南勘测设计研究院有限公司,长沙410014 [2]河海大学水利水电学院,南京210098

出　　处：《长江科学院院报》2025年第3期164-170,共7页Journal of Changjiang River Scientific Research Institute

基　　金：国家自然科学基金项目(52079046)。

摘　　要：为了实现在统计资料匮乏时也能进行重力坝的可靠性分析,构建了仅需要不确定参数上下限的重力坝非概率可靠度评估方法。通过内切椭球模型描述了不确定参数的相关关系,引入比例因子将非概率可靠度计算模型转化为有约束的优化问题,利用Kriging模型对于高度非线性函数的适用性拟合重力坝单元功能函数,在此基础上,通过鲸鱼优化算法进行可靠度的寻优。经过实例分析计算,非概率可靠度计算方法与规范法得出的坝段的抗滑稳定均处于可靠状态,两种方法的结论相同,这验证了基于凸集比例因子和WOA-Kriging模型的重力坝非概率可靠度计算方法可以有效分析重力坝的可靠性。To achieve reliable analysis of gravity dams with limited statistical data,a non-probabilistic reliability assessment method was developed.This method requires only the upper and lower bounds of uncertain parameters.The correlated relationship of these parameters was described using an inscribed ellipsoid model.By introducing a scaling factor,the non-probabilistic reliability calculation model was transformed into a constrained optimization problem.The Kriging model,known for its effectiveness in fitting highly nonlinear functions,was employed to model the functional behavior of gravity dam elements.Additionally,the Whale Optimization Algorithm(WOA)was used for reliability optimization.Case validation confirmed that the non-probabilistic reliability calculation method,based on the convex set scaling factor and the WOA-Kriging model,effectively analyzes the reliability of gravity dams.

关 键 词：重力坝 凸集比例因子 KRIGING模型 有限元模型 非概率可靠度 

分 类 号：TV642.3[水利工程—水利水电工程]