基于LSSVM和GWOPSO算法的桥岸边坡位移反演方法研究

毛伟琦; 李小珍; 王翔; 赵萌

doi:10.19713/j.cnki.43-1423/u.T20231026

您当前的位置：

首页 >

文章列表页 >

基于LSSVM和GWOPSO算法的桥岸边坡位移反演方法研究

桥梁隧道与结构 | 更新时间：2023-12-05

- 基于LSSVM和GWOPSO算法的桥岸边坡位移反演方法研究
- Inversion method of bridge abutment slope displacement based on LSSVM and GWOPSO algorithm
- 铁道科学与工程学报 2023年20卷第11期页码：4299-4310
- 作者机构：
  
  1.西南交通大学土木工程学院，四川成都 610031
  2.桥梁智能与绿色建造全国重点实验室，湖北武汉 430034
  3.中铁大桥局集团有限公司，湖北武汉 430050
- 作者简介：
  
  王翔(1986—)，男，湖南岳阳人，正高级工程师，从事桥梁工程研究；E-mail：wangxiang4143@163.com
- 基金信息：
  
  国家自然科学基金资助项目(52278463;52208505)
- DOI：10.19713/j.cnki.43-1423/u.T20231026
  中图分类号：
扫描看全文
毛伟琦,李小珍,王翔等.基于LSSVM和GWOPSO算法的桥岸边坡位移反演方法研究[J].铁道科学与工程学报,2023,20(11):4299-4310.

MAO Weiqi,LI Xiaozhen,WANG Xiang,et al.Inversion method of bridge abutment slope displacement based on LSSVM and GWOPSO algorithm[J].Journal of Railway Science and Engineering,2023,20(11):4299-4310.
毛伟琦,李小珍,王翔等.基于LSSVM和GWOPSO算法的桥岸边坡位移反演方法研究[J].铁道科学与工程学报,2023,20(11):4299-4310. DOI： 10.19713/j.cnki.43-1423/u.T20231026.

MAO Weiqi,LI Xiaozhen,WANG Xiang,et al.Inversion method of bridge abutment slope displacement based on LSSVM and GWOPSO algorithm[J].Journal of Railway Science and Engineering,2023,20(11):4299-4310. DOI： 10.19713/j.cnki.43-1423/u.T20231026.

摘要

为解决桥岸边坡位移反演中采用传统粒子群算法收敛速度过慢、容易陷入局部最优解的问题，通过引入最差粒子改进策略和最优粒子扰动策略，提出结合灰狼算法改进优化目标方程特性的粒子群优化算法(GWOPSO)。使用最小二乘支持向量机(LSSVM)建立岩土体参数与监测点位移之间的映射关系，根据Flac3D数值模型计算得到训练与检验样本，设置参数初值与粒子个体极值并代入LSSVM进行训练，应用改进的GWOPSO算法搜索LSSVM最优向量机核函数，提高LSSVM的预测精度和适用性。然后，将反分析的目标函数值作为各个粒子的适应度值，利用改进的GWOPSO算法搜索岩土体力学参数的最优组合，进而提出LSSVM和GWOPSO相结合的边坡位移反演方法(LSSVM-GWOPSO算法)，并讨论其可行性。研究结果表明：在标定函数测试下，改进的粒子群优化算法较传统的粒子群优化算法在收敛速度与精度上有大幅度提高，在处理对于单模态和实时性要求较高的多模态优化问题的大型工程项目中能迅速进入局部搜索，并跳出得到全局最优解。与传统的SOPSO算法和SAPPSO算法相比，LSSVM-GWOPSO算法反演精度和反演速度均有较大幅度提高。将研究结果与算法应用到某艰险山区特大桥梁桥岸边坡岩土体参数反演分析，并对边坡后续的位移以及破坏模式进行预测分析，取得较好效果。

Abstract

This study aimed to address the issues of slow convergence and susceptibility to local optima in the displacement inversion of bridge abutment slopes using traditional Particle Swarm Optimization (PSO) algorithms. By incorporating the strategies of the worst particle improvement and the optimal particle perturbation, a Particle Swarm Optimization algorithm (GWOPSO) was proposed, which improved the characteristics of the objective function through the integration of the Grey Wolf Optimization algorithm. Subsequently, the Least Squares Support Vector Machine (LSSVM) was employed to establish the mapping relationship between geotechnical parameters and monitoring point displacements. Training and testing samples were obtained by calculating the numerical model using Flac3D. Initial parameter values and individual particle extrema were set and used for training LSSVM. The improved GWOPSO algorithm was then applied to search for the optimal support vector machine kernel function, thereby enhancing the prediction accuracy and applicability of LSSVM. Subsequently, the objective function value of the inverse analysis was regarded as the fitness value of each particle. The improved GWOPSO algorithm is utilized to search for the optimal combination of geotechnical parameters, leading to the proposed slope displacement inversion method combining LSSVM and GWOPSO (LSSVM-GWOPSO algorithm), and its feasibility is discussed. The research results indicate that, under the calibration function test, the improved Particle Swarm Optimization algorithm demonstrates significantly enhanced convergence speed and accuracy compared to the traditional PSO algorithm. It rapidly enters the local search and escapes to obtain the global optimum, particularly for large-scale engineering projects that require high real-time performance and deal with multi-modal optimization problems. In comparison with the traditional SOPSO algorithm and SAPPSO algorithm, the LSSVM-GWOPSO algorithm exhibits substantial improvements in inversion accuracy and speed. The research findings and algorithms are applied to the inversion analysis of geotechnical parameters for abutment slopes of a major bridge in a challenging mountainous area. The subsequent displacement and failure mode predictions of the slopes yield favorable results.

关键词

桥岸边坡改进PSO算法LSSVM反演分析岩土体参数

Keywords

bridge bank slopeimproved PSO algorithmLSSVMinversion analysisgeotechnical parameters

references

MATSUI T, SAN K C. Finite element slope stability analysis by shear strength reduction technique[J]. Soils and Foundations, 1992, 32(1): 59-70.

MULLER L. The rock slide in the Vajont valley[J]. Rock Mechanicsand Engineering Geology, 1964, 2(1): 148-222.

赵炼恒, 董潇阳, 胡世红, 等. 考虑空间变异性的分层边坡可靠度分析[J]. 铁道科学与工程学报, 2021, 18(12): 3180-3187.

ZHAO Lianheng, DONG Xiaoyang, HU Shihong, et al. Reliability analysis of layered slope considering spatial variability[J]. Journal of Railway Science and Engineering, 2021, 18(12): 3180-3187.

TANG Huiming, HU Xinli, DENG Qinglu, et al. Research on the characteristics and slope deformation regularity of the Badong formation in the three gorges reservoir area[M]// WANG F, LI T. Landslide disaster mitigation in Three Gorges Reservoir, China. Berlin, Heidelberg: Springer, 2009: 87-113.

ZHAO Yue, LUO Jian. Bayesian inverse modeling of large-scale spatial fields on iteratively corrected principal components[J]. Advances in Water Resources, 2021, 151: 103913.

WANG Yankun, TANG Huiming, HUANG Jinsong, et al. A comparative study of different machine learning methods for reservoir landslide displacement prediction[J]. Engineering Geology, 2022, 298: 106544.

温树杰, 赖光甜, 孙自立. 改进的边坡最小势能法及潜在滑动方向研究[J]. 铁道科学与工程学报, 2022, 19(8): 2249-2258.

WEN Shujie, LAI Guangtian, SUN Zili. Study on improved minimum potential energy method of slope and potential sliding direction[J]. Journal of Railway Science and Engineering, 2022, 19(8): 2249-2258.

陈昌富, 李伟, 张嘉睿, 等. 山区公路边坡工程智能分析与设计研究进展[J]. 湖南大学学报(自然科学版), 2022, 49(7): 15-31.

CHEN Changfu, LI Wei, ZHANG Jiarui, et al. State-of-the-art of intelligent analysis and design in slope engineering of highways in mountainous areas[J]. Journal of Hunan University (Natural Sciences), 2022, 49(7): 15-31.

ZHANG J, TANG W H, ZHANG L M. Efficient probabilistic back-analysis of slope stability model parameters[J]. Journal of Geotechnical and Geoenvironmental Engineering, 2010, 136(1): 99-109.

王艳昆. 基于机器学习的滑坡位移区间预测与稳定性分析[D]. 武汉: 中国地质大学, 2020.

WANG Yankun. Interval prediction and stability analysis of landslide displacement based on machine learning[D].Wuhan: China University of Geosciences, 2020.

VAPNIK V N. Setting of the learning problem[M]// The nature of statistical learning theory. New York: Springer, 1995: 15-32.

Guest editorial vapnik-chervonenkis (vc) learning theory and its applications[J]. IEEE Transactions on Neural Networks, 1999, 10(5): 985-987.

陈秋莲, 郑以君, 蒋环宇, 等. 基于神经网络改进粒子群算法的动态路径规划[J]. 华中科技大学学报(自然科学版), 2021, 49(2): 51-55.

CHEN Qiulian, ZHENG Yijun, JIANG Huanyu, et al. Improved particle swarm optimization algorithm based on neural network for dynamic path planning[J]. Journal of Huazhong University of Science and Technology (Natural Science Edition), 2021, 49(2): 51-55.

崔臻, 盛谦. 裂隙岩体等效力学参数结构效应的计算机模拟试验研究[J]. 岩土力学, 2018, 39(10): 3830-3840.

CUI Zhen, SHENG Qian. Numerical modelling of structural effect of equivalent mechanical parameters of fractured rock mass[J]. Rock and Soil Mechanics, 2018, 39(10): 3830-3840.

赵萌. 三峡库区黄土坡滑坡滑带土特性及失稳机理研究[D]. 武汉: 中国地质大学, 2021.

ZHAO Meng. Study on soil characteristics and instability mechanism of loess slope landslide in Three Gorges Reservoir area[D].Wuhan: China University of Geosciences, 2021.

TANG Huiming, LI Changdong, HU Xinli, et al. Deformation response of the Huangtupo landslide to rainfall and the changing levels of the Three Gorges Reservoir[J]. Bulletin of Engineering Geology and the Environment, 2015, 74(3): 933-942.

SUYKENS J A K, VANDEWALLE J. Least Squares support vector machine classifiers[J]. Neural Processing Letters, 1999, 9(3): 293-300.

FAN Yong, PEI Yong, YANG Guangdong, et al. Prediction of blasting vibration velocity peak based on an improved PSO-BP neural network[J]. Journal of Vibration and Shock, 2022, 41(16): 194-203, 302.

ZHANG Yixuan, CUI Ningbo, FENG Yu, et al. Comparison of BP, PSO-BP and statistical models for predicting daily global solar radiation in arid Northwest China[J]. Computers and Electronics in Agriculture, 2019, 164: 104905.

XUE Xinhua, XIAO Ming. Deformation evaluation on surrounding rocks of underground Caverns based on PSO-LSSVM[J]. Tunnelling and Underground Space Technology, 2017, 69: 171-181.

漆祖芳, 姜清辉, 周创兵, 等. 基于v-SVR和MVPSO算法的边坡位移反分析方法及其应用[J]. 岩石力学与工程学报, 2013, 32(6): 1185-1196.

QI Zufang, JIANG Qinghui, ZHOU Chuangbing, et al. A new slope displacement back analysis method based on v-SVR and MVPSO algorithm and its application[J]. Chinese Journal of Rock Mechanics and Engineering, 2013, 32(6): 1185-1196.

李真, 王帆, 王冉珺. 一种结合灰狼算法的粒子群优化算法[J]. 计算机测量与控制, 2021, 29(10): 217-222.

LI Zhen, WANG Fan, WANG Ranjun. An improved particle swarm optimization algorithm[J]. Computer Measurement & Control, 2021, 29(10): 217-222.

PRITHI S, SUMATHI S. Automata based hybrid PSO-GWO algorithm for secured energy efficient optimal routing in wireless sensor network[J]. Wireless Personal Communications, 2021, 117(2): 545-559.

王芝银, 杨志法, 王思敬. 岩石力学位移反演分析回顾及进展[J]. 力学进展, 1998, 28(4): 488-498.

WANG Zhiyin, YANG Zhifa, WANG Sijing. A review on inverse analysis of displacements in rock mechanics[J]. Advances in Mechanics, 1998, 28(4): 488-498.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于监测数据的铁路高边坡力学参数反演及长期稳定性预测