西部寒区隧道冻结泥岩蠕变模型模糊随机辨识  

Fuzzy Random Identification on the Creep Model of Frozen Mudstone in Tunnel Project of Western Cold Region

作　　者：姚亚锋[1] 曹广勇[2] 林键[2] 季京晨 YAO Yafeng;CAO Guangyong;LIN Jian;JI Gingchen(Nantong Vocational University,Nantong,Jiangsu 226001,China;Anhui Jianzhu University,Hefei,Anhui 230022,China)

机构地区：[1]南通职业大学,南通226001 [2]安徽建筑大学,合肥230022

出　　处：《铁道工程学报》2020年第11期83-89,共7页Journal of Railway Engineering Society

基　　金：国家自然科学基金项目(51874005,51374010,51474004);江苏省建设系统科技计划项目(2017ZD062);南通市级科技计划项目(MS12018054);江苏省高校青蓝工程项目。

摘　　要：研究目的:冻结围岩蠕变方程和参数是西部寒区隧道施工安全的关键资料,但已有的研究成果表明蠕变参数主要依靠试验拟合和数理统计的方法获得,未曾兼顾实际岩土工程中冻结岩石的模糊随机性。本文针对西部寒区冻结泥岩分别在-5℃、-10℃、-15℃下进行单轴抗压试验和单轴蠕变试验,揭示冻结泥岩的模糊随机分布。为准确表征西部寒区冻结围岩的不确定性,对传统的粒子群智能算法进行模糊随机改进,利用改进后的算法科学有效地辨识冻结泥岩西原体蠕变模型及其参数。研究结论:(1)西部冻结泥岩在单轴受力状态下,随着温度的降低抗压强度总体呈上升趋势,受温度、含水率、地压等因素的综合影响,其力学特性表现出明显的模糊随机性;(2)相同含水率的西部冻结泥岩,其蠕变过程受冻结温度和应力等级的影响而变化,同时伴有模糊随机性,在较低和中等应力作用条件下,蠕变基本呈稳定性状态;当发生较高应力作用时,蠕变呈加速状态;(3)通过模糊惯性权重对传统的粒子群算法进行模糊随机改进,利用改进后的算法辨识蠕变柔量参数和西原体模型,优化后的模型能更好地表征西部寒区隧道工程中冻结泥岩的蠕变各阶段;(4)改进后的模糊随机粒子群算法比传统粒子群算法变异概率小,在高维复杂问题求解时更容易找到全局最优解,工程适用性更广。Research purposes:The creep equation and parameters of frozen surrounding rock are the key data of tunnel construction safety in western cold region.However,the existing research results show that creep parameters are mainly obtained by means of test fitting and mathematical statistics,but the fuzzy randomness of frozen rock in practical geotechnical engineering has not been considered.In this paper,uniaxial compression test and uniaxial creep test were carried out at-5℃,-10℃and-15℃,respectively,to reveal the fuzzy random distribution of frozen mudstone.In order to accurately characterize the uncertainty of the frozen surrounding rock in the western cold region,the traditional PSO algorithm was improved by fuzzy random method,and the improved algorithm was used to identify the creep model and parameters of the frozen mudstone.Research conclusions:(1)Under uniaxial stress,the compressive strength of frozen mudstone in western China generally increases with the decrease of temperature.Under the comprehensive influence of temperature,water content,ground pressure and other factors,its mechanical properties show obvious fuzzy randomness.(2)The creep process of frozen mudstone with the same water content in the west is affected by freezing temperature and stress level,and is accompanied by fuzzy randomness.Under low and medium stress conditions,creep is basically stable.When the stress is high,the creep is accelerated.(3)The traditional PSO algorithm is improved by fuzzy inertia weight,and the improved algorithm is used to identify creep flexibility parameters and the Nishihara model.The optimized model can better represent the creep stages of frozen mudstone in tunnel engineering in the western cold region.(4)The improved fuzzy random PSO algorithm has less mutation probability than the traditional algorithm,and it is easier to find the global optimal solution when solving complex high-dimensional problems.

关 键 词：冻结泥岩 西原体模型 模糊随机 粒子群算法 蠕变柔量 

分 类 号：TU452[建筑科学—岩土工程]